Content uploaded by Jiří Kvaček
Author content
All content in this area was uploaded by Jiří Kvaček on Apr 30, 2019
Content may be subject to copyright.
Content uploaded by Kleoniki Keklikoglou
Author content
All content in this area was uploaded by Kleoniki Keklikoglou on Apr 30, 2019
Content may be subject to copyright.
Micro-computed tomography for natural history specimens:
a handbook of best practice protocols
Kleoniki KEKLIKOGLOU
1,*, Sarah FAULWETTER
2, Eva CHATZINIKOLAOU
3,
Patricia WILS 4, Jonathan BRECKO
5, Jiří KVAČEK 6,
Brian METSCHER
7 & Christos ARVANITIDIS 8
1,3,8
Institute of Marine Biology, Biotechnology and Aquaculture, Hellenic Centre for Marine Research,
Thalassocosmos, 71003 Heraklion, Crete, Greece.
2
University of Patras, Department of Zoology, Section of Marine Biology, 26504 Patras, Greece.
4 CNRS UMS 2700, Muséum national d’Histoire naturelle, Paris, France.
5
Scientic Heritage Service, Royal Belgian Institute of Natural Sciences, Vautierstraat 29,
B-1000 Brussels, Belgium and Biological Collection and Data Management,
Royal Museum for Central Africa, Leuvensesteenweg 13, B-3080 Tervuren, Belgium.
6 Department of Palaeontology, National Museum Prague,
Václavské náměstí 68, 110 00, Praha 1, Czechia.
7 Department of Theoretical Biology, University of Vienna, Althanstrasse 14, 1090 Vienna, Austria.
* Corresponding author: keklikoglou@hcmr.gr
2
Email: sarahfaulwetter@gmail.com
3
Email: evachatz@hcmr.gr
4
Email: patricia.wils@mnhn.fr
5
Email: jbrecko@naturalsciences.be
6 Email: jiri_kvacek@nm.cz
7 Email: brian.metscher@univie.ac.at
8 Email: arvanitidis@hcmr.gr
1 urn:lsid:zoobank.org:author:5EBBC94A-66D3-45EE-9E38-EDF7CF8B17D1
2 urn:lsid:zoobank.org:author:9BF02566-AF30-47EB-840E-DFC841B6FF84
3 urn:lsid:zoobank.org:author:BBFE2A72-6704-4446-9884-AFB1F6B09A68
4 urn:lsid:zoobank.org:author:CC392691-1414-4624-9965-867BE05CDBAF
5 urn:lsid:zoobank.org:author:7AC9797B-88EB-4844-86B9-C88DF7C06B2E
6 urn:lsid:zoobank.org:author:C2C49FC5-2D9F-4712-A029-9BF2FA9489A9
7 urn:lsid:zoobank.org:author:8777DCF2-AF42-4D51-A640-ABC1770B8572
8 urn:lsid:zoobank.org:author:737F149F-C30C-42EB-A690-5E693AD95427
Abstract. Micro-computed tomography (micro-CT or microtomography) is a non-destructive imaging
technique using X-rays which allows the digitisation of an object in three dimensions. The ability of
micro-CT imaging to visualise both internal and external features of an object, without destroying the
specimen, makes the technique ideal for the digitisation of valuable natural history collections. This
handbook serves as a comprehensive guide to laboratory micro-CT imaging of different types of natural
history specimens, including zoological, botanical, palaeontological and geological samples. The basic
European Journal of Taxonomy 522: 1–55 ISSN 2118-9773
https://doi.org/10.5852/ejt.2019.522 www.europeanjournaloftaxonomy.eu
2019 · Keklikoglou K. et al.
This work is licensed under a Creative Commons Attribution License (CC BY 4.0).
Collection management
urn:lsid:zoobank.org:pub:2B68E2FD-BE81-440B-9A02-470417CC682E
1
principles of the micro-CT technology are presented, as well as protocols, tips and tricks and use cases
for each type of natural history specimen. Finally, data management protocols and a comprehensive list
of institutions with micro-CT facilities, micro-CT manufacturers and relative software are included.
Keywords. Micro-CT, microtomography, museum specimens, 3D visualisation, virtual specimens.
Keklikoglou K., Faulwetter S., Chatzinikolaou E., Wils P., Brecko J., Kvaček J., Metscher B. & Arvanitidis C.
2019. Micro-computed tomography for natural history specimens: a handbook of best practice protocols. European
Journal of Taxonomy 522: 1–55. https://doi.org/10.5852/ejt.2019.522
Table of contents
1. Scope and structure of this handbook .............................................................................................. 3
2. The image acquisition workow ....................................................................................................... 3
3. Protocols for micro-CT image acquisition ....................................................................................... 8
3.1 Pre-scan considerations and specimen preparation ............................................................................ 8
3.1.1 Zoological samples ................................................................................................................... 8
3.1.2 Botanical samples ................................................................................................................... 10
3.1.3 Palaeontological samples .........................................................................................................11
3.1.4 Geological (mineral) samples ................................................................................................. 12
3.2 Scanning containers and scanning mediums ................................................................................ 12
3.3 Scanning process ........................................................................................................................... 15
3.3.1 Calibrating the system ............................................................................................................... 15
3.3.2 Placing the specimen .............................................................................................................. 16
3.3.3 Setting up the detector parameters .......................................................................................... 16
3.4. Reconstruction ............................................................................................................................. 19
3.5 Visualisation and post-processing ................................................................................................. 21
3.5.1 Volume Rendering .................................................................................................................. 22
3.5.2 Isosurface rendering ................................................................................................................ 23
3.5.3 Segmentation .......................................................................................................................... 24
3.6 Troubleshooting ............................................................................................................................ 25
3.6.1 Beam hardening artefacts ....................................................................................................... 25
3.6.2 Ring artefacts .......................................................................................................................... 26
3.6.3 Noise ....................................................................................................................................... 27
3.6.4 Partial volume ......................................................................................................................... 27
3.6.5. Motion artefacts ..................................................................................................................... 27
3.6.6 Metal artefacts ........................................................................................................................ 28
4. Use cases ............................................................................................................................................ 30
4.1 Zoological samples ....................................................................................................................... 30
4.2. Botanical samples ........................................................................................................................ 30
4.3. Palaeontological samples ............................................................................................................. 31
4.4 Geological (mineral) samples ....................................................................................................... 32
5. Data curation .................................................................................................................................... 33
5.1 Documentation .............................................................................................................................. 33
5.2 Data organisation, storage and archival ........................................................................................ 34
5.3 Data dissemination and publication .............................................................................................. 34
5.3.1 Data publication ...................................................................................................................... 34
5.3.2 Tools for outreach and interaction with 3D data ..................................................................... 36
Acknowledgements .............................................................................................................................. 36
References ............................................................................................................................................. 36
Appendix ............................................................................................................................................... 47
European Journal of Taxonomy 522: 1–55 (2019)
2
1. Scope and structure of this handbook
Micro-computed tomography (micro-CT, X-ray computed tomography, high-resolution X-ray computed
tomography, HRXCT/HRCT, high resolution CT, X-ray microscopy) is a non-destructive imaging
technique which allows the creation of high-resolution three-dimensional data. Based on X-ray imaging,
it creates a full virtual representation of both internal and external features of the scanned object. These
resulting 3D models can then be either interactively manipulated on screen (rotation, zoom, virtual
dissection, isolation of features or organs of interest), or an array of sophisticated 3D measurements can
be performed – from simple length and volume measurements to density, porosity, thickness and other
material-related parameters.
While already having been used in geology and palaeontology for decades (e.g., Carlson & Denison
1992; Simons et al. 1997; Rivers et al. 1999; Sutton et al. 2001; Carlson et al. 2003; Rossi et al. 2004;
Burrow et al. 2005; Cnudde et al. 2006; DeVore et al. 2006), in recent years micro-CT has seen a
steep increase of usage in a variety of biological research elds such as taxonomy and systematics,
evolutionary and developmental research and functional morphology (see, e.g., Faulwetter et al. 2013a
and references therein).
The ability of micro-CT imaging to create accurate, virtual representations of both internal and external
features of an object, at micrometer resolution, without destroying the specimen, makes the technique
an interesting tool for the digitisation of valuable natural history collections. Digitisation efforts have
become an important research activity of museums and herbaria, since collections represent a vast
source of biodiversity information which is often underexploited due to the traditionally slow process of
re-visiting physical specimens (Blagoderov et al. 2012). Digitised specimens, however, are available at
the click of a mouse from any internet-enabled computer worldwide, protect the specimen from loss or
damage through handling or shipping and thus have the potential to accelerate taxonomic and systematic
research and allow for large-scale comparative morphological studies (Faulwetter et al. 2013a). Micro-
CT imaging technology may give rise to the elaboration of ‘virtual museums’ or ‘virtual laboratories’
where digital data are shared widely and freely around the world, while the original material is stored
safely (Abel et al. 2011; Keklikoglou et al. 2016). In addition, 3D models created through micro-CT
scanning can be either printed or made available via interactive touch screens to be used for public
display and outreach efforts.
This handbook acts as a comprehensive guide to laboratory micro-CT imaging of different types of
natural history specimens, from geological and palaeontological to zoological and botanical specimens.
First, a general overview of the image acquisition workow is given, presenting the basic principles of
the micro-CT technology. Then, a comprehensive section on best practice protocols follows. For each
of the above categories of natural history specimens, a detailed description of best practices, protocols,
tips and tricks and use cases are given, from specimen preparation to nal use of the resulting models.
However, each specimen is different, and each study has a different scope, so naturally there is no
standard protocol that can be universally applied. The information given in this handbook merely acts as
a starting point. The last section of the main text comprises information on the data management of the
micro-CT datasets, including best practices on metadata, storage and dissemination. Finally, an appendix
includes a glossary which explains the domain-specic terms used throughout the text and additional
useful information, such as lists of institutions with micro-CT facilities, micro-CT manufacturers and
annotated list of software.
2. The image acquisition workow
The basic principle of the micro-CT technique is related to the acquisition of a large set of radiographic
projections (2D images) of an object around a rotation axis. The rotated object is placed between an
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
3
X-ray source and an X-ray detector (Fig. 1) and the free adjustment of the source-object distance (SOD)
and object-detector distance (ODD) allows greater resolution in comparison to clinical CT scanners
(Schambach et al. 2010).
The micro-CT technique depends on X-rays which are high energy electromagnetic radiation ranging
from hundreds of eV to hundreds of keV. X-rays photons are generated by electron beams. The X-ray
source contains an X-ray generator (a vacuum tube) in which electrons are released from a lament (the
cathode) and are highly accelerated by an electric potential difference (Fig. 2). Then, the electronic beam
is focused on a metal target (the anode) and produce X-rays according to two different processes:
The electrons are decelerated by the atomic nucleus of the target and part of their kinetic energy is
converted into an emitted X-ray photon. This phenomenon is called ‘braking radiation’ or Bremsstrahlung.
The output spectrum consists of a continuous spectrum of X-ray energies ranging from 0 to the voltage
of the X-ray source.
Fig. 1. Schematic overview of the image acquisition process. Image by the Hellenic Centre for Marine
Research (HCMR) micro-CT lab.
Fig. 2. Schematic overview of the X-Ray generator.
European Journal of Taxonomy 522: 1–55 (2019)
4
The electrons may collide with target orbital electrons and be ejected from the orbit. Subsequently,
an electron from a higher energy level will replace the ejected electron and the energy left by this
displacement will be transferred to an emitted X-ray photon. The energy of the X-ray photon (uorescent
photon) is the difference between the two energy levels, a characteristic of the target’s material.
The appropriate choice of the target’s material aims to ensure that the energy efciency of the braking
radiation is high (its atomic number is high) and the fusion point is high enough to endure high
power. Tungsten is used as the main material for high power applications. Another typical material is
molybdenum for ner resolution applications. Sutton et al. (2014) mention that a tungsten target source
is considered ideal for palaeontological specimens and a molybdenum target source for specimens which
are included in amber.
Filters (e.g., aluminium, copper, aluminium and copper combined) can be used to prevent the lowest
X-ray photon’s energy reaching the specimen, thus avoiding artefacts during the reconstruction procedure
(beam hardening; Section 3.6). Figure 3 shows a typical spectrum generated by an X-ray generator for
100 kV, along with the spectrum when it is ltered with a 0.5 mm copper plate.
X-rays are attenuated along their paths through the specimen due to three types of interactions:
photoelectric absorption, Rayleigh scattering and Compton scattering. This attenuation is characterised
Fig. 3. The spectrum generated by an X-ray generator at 100kV with and without ltering. Image generated by
the simulation environment https://www.oem-xray-components.siemens.com/x-ray-spectra-simulation.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
5
by a linear coefcient µ (E,Z) in cm-1 that corresponds to the contribution of each type of interaction and
depends on the energy of the incident beam and the atomic number of the material encountered.
The total attenuation of an incident beam passing through a well-dened specimen can be computed as
a sum or an integration of individual attenuations. This property is a key to tomographic reconstruction,
called an inverse problem, as the measurement of the total attenuations on different angles will lead to
knowledge of a discrete set of attenuations (in a voxel grid) along the path through the specimen.
The X-rays photons that have been transmitted through the specimen are then collected on a detection
device. A scintillator screen absorbs the X-ray beam and re-emits it in the form of light. This light may
be captured by CCD or CMOS cameras, digitised by a photodiode array in a at-panel detector. The
choice of a detector is usually a trade-off between its pixel resolution and its eld of view.
The resulting image is a radiograph (projection image) whose pixel values are the X-ray transmission
as measured by the detector, usually mapped to 16-bit gray values (ranging from 0 to 65535). This type
of acquisition is called absorption-contrast. When displayed as a positive image, the darkest parts of an
X-ray image are the most absorbing ones and the lightest parts are those with the lowest absorption (e.g.,
air) (Fig. 4). The reconstructed tomographic image consists of voxels whose values correspond to the
X-ray attenuation at each point in the sample.
The aforementioned projection images are reconstructed into cross-section images using specic
reconstruction software. A reconstruction algorithm is run to get a volumetric representation of the
density of the specimen, including its inner features. The most common reconstruction software
use ltered back projection algorithms for the recovery of the attenuation maps of the radiographs.
Fig. 4. Example of the projection images resulting from the scanning process. Image by HCMR micro-
CT lab.
European Journal of Taxonomy 522: 1–55 (2019)
6
Specically, projection images, which are taken from every angle of the sample, results in sinograms
which represents the aforementioned attenuation maps (Betz et al. 2007). The cross-section images
(slices) are created by these sinograms using back-lter algorithms. Each projection image is smeared
back across the reconstructed image and creates the back-projection images which transmit the measured
sinogram back into the image space along the projection paths. The back-projection image is a blurred
cross-sectional image. This blurring effect can be moderated using mathematical lters (Sutton et al.
2014) and the algorithms that use the combination of back projection and ltering are known as ltered
back-projection algorithms. The combination of the back-projection images will localise the position of
the sample. As the number of projections increase, the position and shape of the object becomes more
dened (Fig. 5).
The reconstruction procedure results in a 3D image where each voxel codes the local density of the
specimen. It is usually exported as a stack of 2D images in a given orientation (Fig. 6). The micro-
CT dataset can then be processed with dedicated software for 3D visualisation (either through
volume rendering or isosurface rendering, see Section 3.5) or further analysis (see Tables 8 and 9 for
comprehensive lists of software). The rendered images can be also used for the creation of 3D videos
which are an excellent tool for sharing a preview of a micro-CT dataset (Abel et al. 2012).
Fig. 5. Schematic overview of the reconstruction procedure. Image by HCMR micro-CT lab.
Fig. 6. Data processing from a stack of 2D images to a 3D model. Image by MNHN.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
7
3. Protocols for micro-CT image acquisition
3.1. Pre-scan considerations and specimen preparation
The selection of micro-CT as a technique for specimen visualisation depends on the aim of the specic
study. The limitations of the method must be considered, e.g., colour cannot be represented and
structures below a certain size cannot be detected. An additional consideration when working with
museum specimens is to ensure that the museum curator will allow imaging with micro-CT (including
contrast enhancement through staining, if needed).
The rst and major step to obtain good results is to achieve a sound contrast difference between the
specimen and its surrounding medium. Contrast enhancement agents can be used to improve the quality
and clarity of the scanning result: 1) when the specimen has inadequate contrast, 2) when extreme
attenuation differences between soft and hard tissue need to be reduced (setting of the contrast level in
the scan becomes easier), 3) to segment tissues or organs more easily, and 4) to identify specic target
features of the specimen.
Rehydration or dehydration procedures may be needed before staining, depending on the preservation
medium of the specimen, or after staining and on the medium in which the specimen will be scanned.
3.1.1. Zoological samples
Dense material, such as bones and other calcied tissues, usually does not require any specic
preparation. In soft tissue specimens, where visualisation can be difcult due to the low X-ray absorption
of unmineralised tissues (Metscher 2009a; Gignac & Kley 2014), contrast enhancement agents are
commonly applied.
A series of the most common contrast agents used in zoological samples is presented in Table 1. In
general, contrast agents with a high atomic number are more efcient since they result in an increased
X-ray absorption (Pauwels et al. 2013). However, the selection of an appropriate and efcient contrast
enhancement agent is a combination of several considerations such as the type of the target tissue (see
Table 1), the medium in which the specimen was xed and stored (Metscher 2009a), the acidity and the
penetration rate of the contrast agent (Metscher 2009b; Pauwels et al. 2013; Paterson et al. 2014), as
well as the price and toxicity of the staining used. Contrast agents with low penetration rate are more
effective when used in relatively small specimens with a sample size of only a few cm3 (Pauwels et al.
2013). Staining larger specimens may require longer staining time, as the penetration rate in larger
samples may be slower. Contrast agents dissolved in buffered formalin could prevent potential tissue
deterioration due to the long staining period for large specimens (Li et al. 2015).
Staining of fresh samples will usually give optimal results. However, the results of the staining can be
inuenced by certain treatments, such as freezing prior to xation and the xation process (Gignac et al.
2016). If possible, long-term storage in ethanol or unbuffered formalin between xation and staining
should be avoided, as this may affect the morphology of the specimen or alter tissue characteristics
which in turn affect staining properties (e.g., iodine stains bind to lipids, which can be dissolved in
alcohols (Gignac et al. 2016), and unbuffered formalin or other acidic liquids may decalcify tissues).
However, this is not always possible for museum specimens, and good results have been also achieved
with museum specimens stored for years and decades. These ner details of xative, storage medium,
tissue characteristics and staining properties are still insufciently known and will require further studies
in the future.
Micro-CT scanning is a powerful visualisation method with several advantages, but might not be
appropriate for all kinds of specimens. Some contrast enhancement agents are acidic (e.g., PTA, PMA,
FeCl3) and when used in high concentrations they may dissolve calcied tissues, such as bones, and
destroy the specimen structure (Pauwels et al. 2013). Therefore, in cases where calcied structures are
European Journal of Taxonomy 522: 1–55 (2019)
8
Contrast agent Tissue Limitations Dissolved in References
Osmium (osmium
tetroxide, OsO4)
membrane lipids, proteins
and nucleic acids
toxic, volatile, expensive,
does not work for tissues
preserved in ethanol, slow
penetration
phosphate
buffer
Metscher 2009b;
Kamenz &
Weidemann 2009;
Faraj et al. 2009
Phosphotungstic acid
(PTA)
various proteins (overall
structure), connective
tissue (collagen), muscle
cartilage matrix does
not stain strongly, slow
penetration, potential
dissolution of calcied
tissues if high concentration
of contrast agent is used
water or
ethanol-
methanol
Metscher 2009b;
Pauwels et al. 2013
Phosphomolybdic acid
(PMA)
cartilage structures,
brous collagen tissue
slow penetration, potential
dissolution of bone structure
if high concentration of
contrast agent is used
water
Golding & Jones
2007; Pauwels et al.
2013
Iodine overall structure of soft
tissues
may overstain some
mineralised tissues
ethanol or
methanol
Metscher 2009b;
Gignac et al. 2016
Iodine potassium
iodide (IKI)
overall structure of soft
tissues
may overstain some
mineralised tissues water Metscher 2009b
Iodine-based buffered
formalin
overall structure of soft
tissues
tissue shrinkage at high
concentrations
neutral buffered
formalin
Li et al. 2015;
Bribiesca-Contreras
& Sellers 2017
Lugol’s iodine (I2KI) glycogen, lipids
(carbohydrates)
tissue shrinkage at high
concentrations, may
overstain some mineralised
tissues
water Gignac & Kley 2014
Gold (Bodian
impregnation) neuron (neuropils)
penetration at 100 μm,
complex creation of the
staining solution
complex, see
Mizutani &
Suzuki (2012)
Mizutani & Suzuki
2012; Mizutani et al.
2007
Silver (Golgi silver
impregnation)
neuron, cerebral
cortex, proteins in
polyacrylamide gels, lung
tissue
visualisation only 10% of
neuron, complex creation of
the staining solution
complex, see
Mizutani &
Suzuki (2012)
Mizutani & Suzuki
2012; Paterson et al.
2014
Platinum neuron (substitute of gold) complex creation of the
staining solution
complex, see
Mizutani et al.
(2008a, 2008b)
Mizutani et al. 2008a,
2008b
Mercury(II) chloride
(HgCl2)
cerebral cortex tissue,
muscle, brous collagen
tissue, ligaments, large
blood vessels
slow penetration
complex, see
Mizutani et al.
(2009)
Pauwels et al. 2013;
Mizutani et al. 2009
Lead muscle, brous collagen
tissue
precipitations in and around
structures water
Faraj et al. 2009;
Pauwels et al.
2013; Kamenz &
Weidemann 2009
Barium based (e.g.,
BaSO4, BaCl2)
biolm in porous media,
brous collagen tissue
precipitations in and around
bone structure water Davit et al. 2011;
Pauwels et al. 2013
Iron based (e.g. FeCl3) nucleic acids, proteins slow penetration, dissolution
of bone structure water Pauwels et al. 2013;
Paterson et al. 2014
Hexamethyldisilizane
(HMDS)
removes water from
tissues, increasing clarity
of boundaries between air
and tissue
possible internal tissue
damage, renders specimens
fragile, may react with metal
stains previously used on
specimen
air
Alba-Tercedor &
Sánchez-Tocino 2011;
Paterson et al. 2014
Table 1 (continued on next page). Overview of the most common contrast agents for soft zoological
tissues.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
9
included in the specimen tissue the minimum, but still effective concentration of acidic agents needs
to be identied prior to staining, or alternatively other non-acidic agents should be used for contrast
enhancement.
Contrast agents dissolved in ethanol (e.g., PTA, iodine) may cause shrinkage when used on specimens
which are not stored in ethanol. In such cases a water based contrast enhancement agent may be more
appropriate and safe to use; otherwise a gradual dehydration procedure needs to be followed prior to
staining. Shrinkage due to desiccation may destroy the sample and in addition it can create movement
artefacts during scanning (Johnson et al. 2011). Samples can also be critical point dried, freeze-dried or
chemically dried with Hexamethyldisilazane (HMDS) and then scanned without any surrounding liquid
medium in order to increase the contrast between tissues. Generally, these methods can dry the samples
without inducing morphological changes to the tissues, although they might render the sample fragile or
cause moderate shrinkage (Faulwetter et al. 2013b; Pauwels et al. 2013; Krings et al. 2017). The drying
can be performed both on stained and unstained samples, but care needs to be taken with HMDS, which
may react chemically if applied in combination with certain stains (e.g., silver stains, see Paterson et al.
2014).
Specimens from museum collections usually need to remain completely intact, and thus any alterations
that might be caused by the staining procedure need to be taken into account. Besides the alterations
previously mentioned (decalcication, shrinkage) the removal of the contrast enhancement agents after
scanning is an additional important consideration before using this method on museum specimens. The
stability of the stain might depend on the xation or preservation medium, the age of the specimen, and
the type of tissue (e.g., chitin, muscles, calcied tissue) (Schmidbaur et al. 2015). Iodine staining was
successfully removed from insects (Alba-Tercedor 2012) and from millipedes (Akkari et al. 2015) by
re-immersion into 70% ethanol, and from polychaete tissues using 96% ethanol for 48 hours, while PTA
stain was removed using NaOH for 6 hours (Schmidbaur et al. 2015). However, treatment of iodine
stained tissues with sodium thiosulfate and of PTA stained tissues with 0.1 M phosphate buffer (pH 8.9)
destained the samples even more than their initial unstained status, thus indicating an actual alteration of
the tissues characteristics (Schmidbaur et al. 2015). Gignac et al. (2016) also indicated that destaining
does not really restore a specimen to its original chemical state, e.g., when iodine stained tissues are
treated with sodium thiosulfate, the iodine is transformed to iodide, which is colorless and remains in
the tissues.
3.1.2. Botanical samples
Micro-CT scanning of botanical specimens is usually restricted to non-pressed specimens, i.e., those
that possess a certain three-dimensional structure. Technically, pressed herbarium specimens can also
Contrast agent Tissue Limitations Dissolved in References
Zinc chloride (ZnCl2)muscle, brous collagen
tissue slow penetration water Pauwels et al. 2013
Ammonium
molybdate tetrahydrate
(NH4)2MoO4
muscle, brous collagen
tissue slow penetration water Pauwels et al. 2013
Sodium tungstate
(Na2WO4)
muscle, brous collagen
tissue unknown water Pauwels et al. 2013;
Kim et al. 2015
Gallocyanin-
chromalum nucleic acids low overall contrast water
Presnell &
Schreibman 1997;
Metscher 2009b
Table 1 (continued).
European Journal of Taxonomy 522: 1–55 (2019)
10
be scanned under certain circumstances, but the results will likely be of limited research value. Suitable
botanical specimens include soft tissues (e.g., owers, leaves, buds, fruits) and hard tissues (e.g., stems,
twigs, roots, nuts). Generally, ligneous hard tissues are more easy to scan than soft tissues, as they do
not dry out easily and provide a good contrast due to the higher density of their secondary cell walls.
Plant tissues can often be scanned without any need for xing, preservation of sample, or application
of contrast agents. Fruits, nuts, thick roots and wooden structures and even owers, if provided with a
liquid environment around the stem (van der Niet et al. 2010) can often simply be scanned fresh without
any further preparation. However, if a high resolution of these tissues is required, smaller pieces may
be cut from the original sample to decrease the camera-sample distance. Such smaller samples dry
out faster and thus may require additional means to prevent dehydration such as wrapping in plastic
or Paralm®, scanning in a sealed container, or coating the specimen with additional materials (e.g.,
Korte & Porembski 2011).
Soft tissues are often transparent to X-rays and may require the use of contrast agents, as well as
additional preparation to prevent shrinkage during scanning due to desiccation (Stuppy et al. 2003;
Leroux et al. 2009). If the sample needs to be xed and/or stained, a variety of solutions are available.
Common xatives for botanical samples are FAA (formalin–acetic acid–alcohol), formaldehyde, or
ethanol (e.g., Leroux et al. 2009; Staedler et al. 2013). However, ethanol has been shown to induce
shrinkage in plant tissues and might not be appropriate for all types of studies, e.g., vascular cylinders
might be compressed or ruptured if ethanol is used (Leroux et al. 2009).
Samples xed in a liquid substance will usually require being scanned in a liquid environment as well,
either fully submerged or sealed in a small chamber to prevent drying out. Alternatively, samples can be
embedded in agarose or paraplast, but as this will introduce noise and reduce contrast these media are
only recommended for samples that are either naturally dense or have been treated with heavy-metal
stains (see Table 2).
Table 2 (continued on next page). Overview of the most common contrast agents (heavy metals) for soft
botanical tissues.
Contrast agents Effects and limitations References
Uranyl acetate toxic, only slight increase of contrast values compared to other
stains
Leroux et al. 2009;
Staedler et al. 2013
Iodine no noticeable contrast increase
Korte & Porembski
2011; Staedler et al.
2013
Potassium
permanganate
permanganate caused visible damage to samples as soon as after 2d
inltration (Fig. 5F), and inltration for 8 days usually resulted in
total sample loss; occasionally increased the contrast of only a part
of the sample, leaving other parts unchanged
Staedler et al. 2013
Lugol’s solution causes visible sample damage after several weeks of inltration Staedler et al. 2013
Phosphotungstate
(PTA)
highest contrast increase on the more cytoplasm- and protein-rich
tissues (ovules, ovary wall and pollen) Staedler et al. 2013
Lead citrate
work best on vacuolated tissues (petals, sepals and laments);
highest contrast increase on the more cytoplasm- and protein-rich
tissues (ovules, ovary wall and pollen); precipitates in presence
of carbon dioxide in the form of lead carbonate crystals [38] that
accumulate on the surface of the sample
Staedler et al. 2013
Bismuth tartrate
work best on vacuolated tissues (petals, sepals and laments);
highest contrast increase on the more cytoplasm- and protein-rich
tissues (ovules, ovary wall and pollen); rendered the samples very
delicate and easy to damage
Staedler et al. 2013
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
11
A variety of contrast enhancing heavy metal stains has been tested on plant tissues, to varying levels
of success and with different tissue specicities. A thorough comparative study has been performed by
Staedler et al. (2013). Table 2 summarises the application areas and effects of various heavy metal stains.
3.1.3. Palaeontological samples
Palaeontological specimens (i.e., fossils) may need to be isolated from rock matrices before scanning in
cases where the specimen and the matrix show a similar X-ray absorption. Fossils can be extracted from
their matrix mechanically by washing, wet sieving and the use of several tools such as needles, knives
and chisels (Green 2001; Sutton 2008) or chemically, depending on the chemical composition of the
surrounding matrix. For example, fossils embedded in calcareous rocks can be isolated using sulphuric
acid (Vodrážka 2009), phosphatic fossils can be extracted using acetic acid (Jeppsson et al. 1999) and
plant mesofossils in clay or mud stones can be extracted in water with sodium carbonate, potassium
hydroxide or hydrogen peroxide (Wellman & Axe 1999). The maceration procedures of palaeontological
specimens usually involve physical breakdown, removal of calcareous material, removal of siliceous
material, removal of other inorganic material, oxidation, sieving, cleaning and concentrating the organic
rich residue (Green 2001). A detailed manual for extraction techniques in palaeontology can be found in
Green (2001). However, potential damage of the specimen using chemical extraction methods must be
considered (Sutton 2008). If there is no possibility of extraction of the fossil and the contrast between
the specimen and the matrix is too low, a synchrotron phase-contrast imaging may be a better solution
for the visualisation of such specimens (Sutton et al. 2014). Specimens in amber can be scanned directly
without any particular preparation.
3.1.4. Geological (mineral) samples
Strictly speaking, the only preparation that is absolutely necessary for scanning geological specimens is
to ensure that the object ts inside the eld of view and that it does not move during the scan (Ketcham &
Carlson 2001). Since the full scan eld is a cylinder, it is suggested to scan an object of cylindrical
geometry, either by using a coring drill to obtain a cylindrical sample of the geological material being
scanned, or by packing the object in a cylindrical container with either X-ray-transparent ller or with
material of similar density (Ketcham & Carlson 2001).
For some applications the sample can also be treated to enhance the contrast between the different
structures. Examples include injecting soils and reservoir rocks with NaI-laced uids to reveal uid-ow
characteristics (Wellington & Vinegar 1987), injecting sandstones with Woods metal to map out the ne-
scale permeability, immersion in caesium chloride to visualise connected porosity of crystalline rocks
(Kuva et al. 2018) and soaking samples in water to emphasize areas of differing permeability, which can
help to reveal fossils (Zinsmeister & De Nooyer 1996).
3.2. Scanning containers and scanning mediums
At the end of the staining procedure, the sample is removed from the solution and washed in distilled
water or ethanol (depending on the contrast agent solvent). However, Staedler et al. (2018) immediately
xed the plant specimens with 1% PTA in formalin–acetic acid–alcohol (PTA/FAA) without washing
Contrast agents Effects and limitations References
Osmium tetroxide
work best on vacuolated tissues (petals, sepals and laments);
highest contrast increase on the more cytoplasm- and protein-rich
tissues (ovules, ovary wall and pollen); poor penetration for en-bloc
inltration → best on open and thin material (open buds, tissues
only a few cells thick)
Staedler et al. 2013
Iron diamine samples could not be detected (no increase in contrast) Staedler et al. 2013
Table 2 (continued).
European Journal of Taxonomy 522: 1–55 (2019)
12
the samples with ethanol. If specimens are stained or naturally dense they can be scanned in liquid or gel
media (e.g., water, ethanol, agarose). If they are to be scanned in air, excess uid should be blotted away
in order to prevent motion artefacts during scanning due to the accumulation of uid on the bottom of
the container (Gignac et al. 2016). Then, the specimen is placed in a sample holder and stabilised in a
vertical position for the scanning procedure. The choice of the appropriate sample holder depends on the
sample size, the morphology of the specimen and the material of the sample holder.
Micro-CT companies usually provide a range of several sample holders to the users. It is essential that
specimens do not move, settle or wobble during scanning; even a small shift can ruin the image data and
necessitate a re-scan (Sutton et al. 2014). Nevertheless, users often create new sample holders according
to their needs in order to prevent movement of the specimen during scanning. Glass containers are
rarely used since they have a high X-ray absorption; however, glass presents a higher contrast with
the background compared to the plastic container, therefore allowing more accurate delineation of the
region of interest which corresponds to the inner part of the container (Paquit et al. 2011). Materials such
as polypropylene, styrofoam and orist foam have low X-ray absorption, making them ideal as sample
holders. It is also quite common to use pins and clips in order to stabilise the specimen; however, care
must be taken in order not to damage museum samples. Sealable plastic bags from which air has been
pressed out can be used to prevent drying of specimens (Gignac et al. 2016). Small specimens can be
placed in test tubes, aliquot tubes, plastic straws or micropipette tips (200, 1000, 5000 μL) (Metscher
2011; Alba-Tercedor 2012; Staedler et al. 2013; Gignac et al. 2016) (Figs 7, 8B–D). The bottom end
of pipette tips can be cut off in order to drag down the samples and to stabilise them by entering a
preparation needle (Staedler et al. 2018). Parafn wax can be used in the lower end of the pipette tip as
a seal and to stabilise the sample, while Paralm® can be used to seal the upper end of the tip (Staedler
et al. 2013; Fig 8C). The lower end of pipette tips can also be heat-sealed if they need to be lled with
a liquid medium surrounding the specimen. Batch scanning can help to scan large numbers of small
samples at the same time and thus decrease overall scanning time: individual samples mounted in pipette
Fig. 7. Setup of scanning containers. Images by HCMR micro-CT lab.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
13
tips can be mounted in a 1 ml syringe tube and stabilised with resin or epoxy glue (Staedler et al. 2013,
2018; Fig. 8).
Specimens can be submerged in uid (e.g., formalin, ethanol or water) during scanning to prevent
desiccation (Gignac et al. 2016). Ethanol is less dense than water and provides greater tissue contrast
in comparison to water (Metscher 2009a). Scanning small specimens within a liquid medium prevents
them from getting stuck on the container’s walls. Care must be taken to remove bubbles from the
liquid medium in the sample holder as they can create a blurring effect (Metscher 2009a). Agarose can
also be used as a scanning medium for small specimens (Metscher 2011). The use of air as a scanning
medium gives excellent results for unstained or dried specimens or for studying the internal structure
of specimens. However, scanning in air is not recommended when studying the external morphology
of specimens previously immersed in liquid, since the small amount of liquid that will remain on the
external surface will be visible as an additional layer in the reconstruction images (Fig. 9).
Fig. 8. Sample mounting techniques for plant specimens of different sizes. A. Large samples (>10 mm).
B–D. Medium-sized samples (1–10 mm). E–F. Small samples (<1 mm). Image from Staedler et al.
2013, reproduced under a CC-BY license.
Fig. 9. External morphology of the polychaete Lumbrineris latreilli Audouin & Milne Edwards, 1834.
Specimen had been preserved in ethanol and was subsequently blotted dry on a tissue and scanned in air.
The remaining ethanol (shown by arrows) can be observed clinging to the anterior end of the polychaete.
Image by HCMR micro-CT lab, CC-BY Sarah Faulwetter.
European Journal of Taxonomy 522: 1–55 (2019)
14
Large plant samples (>10 mm) can be mounted in acrylic foam and scanned in a solvent atmosphere,
whereas medium-sized plant samples are best scanned immersed in the solvent (Staedler et al. 2013;
Fig. 8A–B). Red and brown algae are easily scanned after having being dried in air. Scanning of the
more lamentous green algae and seagrasses within a liquid medium creates artefacts, because their
leaves have a low X-ray absorption. Personal experiments showed that staining with PTA or iodine
could not improve the resolution. Chemical drying of lamentous algae (e.g., HMDS) and scanning in
air can have satisfactory results, but due to shrinkage this method is not suggested when aiming to study
the internal leaf structure.
Palaeontological specimens are usually scanned in air. However, a celluloid lm in organic water-soluble
glue is suitable as a scanning medium for isolated microfossils (Görög et al. 2012). Isolated specimens
can be also xed on plastic holders by nail polish (or any other reversible xing matter). For combined
SEM and micro-CT studies SEM holders can be used. However, xation of fossils must be on a thick
layer of nail polish to avoid contact or close placement of the observed fossil and the metallic holder,
which can occlude the view of the sample. Larger specimens are xed in specially prepared holders
made of plastic or polystyrene. The main concern is to prevent potential movements of the fossil during
the radiography process. Sutton et al. (2014) mentioned that different X-ray penetration between the
different axes of a palaeontological sample could cause noise and artefacts and they suggest to bury the
specimens in a substance such as our for low density samples and sand for more dense samples.
3.3. Scanning process
Scanning settings differ in terms of voltage, exposure time, magnication and resolution depending on
the scope of study, the material of the specimen, the specimen size and the instrument used. A careful
balance between all the scanning parameters is necessary to ensure an ideal result and a best practice
protocol.
3.3.1. Calibrating the system
In order to achieve micrometer precision, the incident beam needs to be thin, focused and stable during the
acquisition. This calibration of the source includes a software-driven warm-up and focusing. Depending
on the scanning system, the user should set the main parameters, including the metal of the source target
(if the instrument has this option), the amount of ltration, the source intensity (in mA) and the source
voltage (in kV). The source voltage should be set up in a way that the beam will have sufcient energy to
penetrate the specimen and reach the detector. Furthermore, the source intensity should not be too high
as the detector may be saturated. On a projection image, the dynamic range (max-min gray value) has to
be maximised. A good contrast on the set of projections leads to a good contrast on the nal CT image.
Too much transmission will reduce the contrast between different densities, while a low transmission
will increase the noise level in the images. The contrast should be checked over a complete rotation of
the specimen. Adjustment of lter and voltage settings should aim at a minimum transmission between
10 and 50%.
Scanning of unstained soft tissues usually does not require the use of lters, as they are characterised
by a high X-ray transmission. An exception could be a sample which is characterised by a combination
of soft and dense structures (e.g., a vertebrate organism), where the use of a lter may be helpful.
Dense structures, such as bones, shells and other calcied structures, are characterised by a high X-ray
absorption, as they contain elements with high atomic number (Schambach et al. 2010). Such structures
appear dark on the images and they have low X-ray transmission, so the use of lters during scanning
is necessary. These lters can be made of aluminium, copper or a combination of the two. Filters reduce
artefacts caused from beam hardening effects (Meganck et al. 2009; Abel et al. 2012), while the spread
of the X-ray energy distribution is reduced (see Section 3.6). However, the use of lters shifts the
grayscale values downwards, resulting in less contrast between tissues (Meganck et al. 2009). Contrast
between tissues can be achieved by decreasing the voltage and in addition, the use lters in cases of
dense structures is recommended for artefact reduction.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
15
3.3.2. Placing the specimen
The specimen is placed on a rotating platform between the X-ray source and the detector. It has to
be centred vertically and horizontally along the detector. A high-precision rotating mount may help
to centre small specimens. When acquiring a specic region of interest, this region has to be centred
instead of the whole specimen. The specimen must not touch either the source or the detector at any time
during a complete rotation.
The magnication depends on the distance between the source, the specimen and the detector. The
nal image voxel size V results from the equation V = p × ss/sd , where p is the detector pixel size, ss
the source to specimen distance and sd the source to detector distance. The distances ss and sd may
be automatically set by the system or may need a calibration. When the positions of the detector and
the rotating platform are set, the calibration usually consists in acquiring a set of radiographs of a
calibration object placed on the rotating platform. A greater magnication and resolution could decrease
beam-hardening effects (see Chapter 3.6), but will increase scanning duration. A big specimen size can
prevent achieving high magnication and resolution while a smaller eld of view is required for the
identication of the smallest structures (Dixon et al. 2018). For this reason, micro-CT datasets acquired
at different resolutions can be combined in order to provide more information (Dixon et al. 2018 and
examples therein).
3.3.3. Setting up the detector parameters
The calibration of the detector is software-driven (but may need regular initiation by the user) and
includes the acquisition of two types of images:
the dark eld image is the resulting image when no X-radiation is emitted. This signal comes from the
dark current in the photodiodes of the detector. This image is an offset that will be subtracted from every
radiograph.
the open eld images are the response of the detector pixels to the incident beam when no specimen is
placed in the system. The software may require a couple of images at different beam intensities or only
one for the maximum beam intensity.
In case some pixels are defective, their response will strongly differ from that of their neighbouring
ones. When they are detected, a defective map is built and these pixels are ignored during an acquisition.
Their values are replaced by an interpolation of the values of non defective neighbouring pixels. The
defect map should be computed once a month.
Several parameters can usually be changed to inuence the imaging process; however, not all scanner
models offer the same options. A few important parameters are listed below.
The exposure time E relies on the same principle as the photographic exposure time. It controls the
amount of time (in seconds) during which the X-rays will be captured. The detector should collect
enough photons at each angle to ensure a good contrast on the radiograph. The effective dynamic range
of the image is proportional to the exposure time if it is not saturated, and hence the short exposure
results in the low signal-to-noise ratio. High density or thick specimens (e.g., fossils) will need a longer
exposure time because fewer photons will be transmitted through them. However, excessively high
exposure times may saturate the detector panel (i.e., raise brightness above its maximum measurable
threshold). The scanning duration is obviously longer when the exposure times are longer. The use of
higher voltage can result in a decrease of the exposure time.
A way to ensure a sufcient contrast without ending up with long scans is to use a binning parameter.
Instead of having single cells (or pixels) collecting X-rays on the detector, the signal is acquired using
‘bins’ of 4 adjacent pixels (for a 2×2 binning). The X-ray ux per (binned) pixel is four times higher and
the exposure time can be reduced accordingly. The pixel size and the resulting voxel size are twice as
European Journal of Taxonomy 522: 1–55 (2019)
16
large in linear dimension (and the resulting volume image is ⅛ as large). Binned acquisition results in a
well-contrasted fast scan, but at the price of lower resolution.
An averaging parameter A (frame averaging) may be set to improve the image quality. A set of images
will be acquired for every angle and only the mean image will be recorded. A higher number of frames
increases the signal-to-noise ratio. Therefore, it is usually recommended to increase the frame averaging
for high dense samples when the signal-to-noise ratio is too low.
Depending on the detector employed, a pausing parameter P can be used to prevent an afterimage effect.
The detector photodiodes need a few ms to get cleared of the image. If the acquisition is too fast, residual
information from the previous image can appear on the next image. The pausing parameter therefore
helps to fully discard the signal from the previous image. It also ensures that the specimen is perfectly
still after the rotation.
The number N of radiographic images needed to perform a reconstruction can be estimated by measuring
the maximum width W of the projected specimen on a radiograph (in pixels) using the formula N =
π × W (for a 360° scan; Fig. 10). Note that to determine the maximum width the specimen should
be rotated, as irregularly shaped specimens may have different widths at different angles. Acquiring
less than N projection images will provide an incomplete dataset for the reconstruction algorithm. The
reconstruction will be still possible, but its quality will be degraded.
The total scanning time T (in seconds) can be computed with the following formula: T = N × (P + (E × A)).
Other scanning settings include the selection of the rotation step, random movement and 180/360° scan.
The selection of an increased rotation step is useful when the scanning duration needs to be reduced but
it can result in images of reduced quality and increased noise. Random movement can be activated to
reduce ring artefacts, but should not be used when the position of the samples is not secured or when
the pixel sizes are very small. The full 360° rotation is selected for scans where the sample consists of a
combination of high dense materials inside low dense materials and helps to avoid depletion artefacts.
The half (180°) rotation can be used when is it necessary to reduce the scanning time. The simultaneous
scan of multiple samples (batch scan) can be used to reduce the overall scanning duration. A short scan
duration is benecial when specimens are scanned in air and therefore dehydration and shrinkage need
to be avoided.
Fig. 10. Measuring the maximum width W (in pixels) of the projected specimen (as the distance from
the rotation axis - dotted line - to the farthest end of the sample) to calculate the number of radiographs
needed. This measurement is done for the angular position of the rotating platform where the projected
specimen is the widest. For a complete rotation, the projected specimen would stay within the limits of
the rectangle. Photo by MNHN.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
17
Examples regarding the scanning parameters of zoological, botanical, geological and palaeontological
samples are included in Tables 3–6.
Specimen Voltage
(kV)
Current
(μΑ) Filter System Source Reference
Alligator mississippiensis
(vertebrate) 150–200 130–145 copper GE phoenix
v|tome|x s240 molybdenum Gignac & Kley 2014
Dromaius novaehollandiae
(vertebrate) 130–180 145–190 copper GE phoenix
v|tome|x s240 molybdenum Gignac & Kley 2014
Polychaeta (invertebrates) 60 167 none SkyScan 1172 tungsten Faulwetter et al. 2013a
Apporectodea caliginosa
Apporectodea trapezoides
(anterior part – invertebrates)
65–70 100 none Skyscan 1173 tungsten Fernández et al. 2014
Ommatoiulus avatar
(invertebrate) 60 67 none Zeiss/Xradia
MicroXCT-200 tungsten Akkari et al. 2015
Snake embryos (vertebrates) 40 100 none Zeiss/Xradia
MicroXCT-200 tungsten van Soldt et al. 2015
Table 3. Examples of scanning parameters for zoological samples.
Specimen Voltage
(kV)
Current
(μΑ) Filter System Source Reference
root samples of
Asplenium theciferum 50 unknown none
in-house nano-CT
at University of
Ghent
unknown Leroux et al. 2009
Xylem of Laurus 50 275 none Nanotom 180 XS;
GE unknown Cochard et al. 2015
macadamia nuts-in-shell 60 unknown unknown SkyScan 1172 tungsten Plougonven et al.
2012
Flower of Bulbophyllum
bicoloratum 50 100 unknown Xradia Micro-
XCT-200 tungsten Gamisch et al. 2013
Table 4. Examples of scanning parameters for botanical samples.
Specimen Voltage (kV) Current
(μA) Filter System Source Reference
Meteorite 120–160 60–100 unknown
Nikon
Metrology HMX ST 225 tungsten Hezel et al. 2013
Meteorite 180 120 unknown
Nikon
Metrology HMX ST 225 tungsten Needham et al. 2013
Meteorite 200 160 unknown Nikon
Metrology HMX ST 225 tungsten Grifn et al. 2012
Crystalline rock 100 80 aluminium
and copper SkyScan 1172 tungsten Kuva et al. 2018
Table 5. Examples of scanning parameters for geological samples.
European Journal of Taxonomy 522: 1–55 (2019)
18
3.4. Reconstruction
The reconstruction workow varies among the different scanning systems as each system usually has
its own software and most reconstruction steps are automated (Sutton et al. 2014). However, the setting
of some reconstruction parameters is necessary and depends on the scope of the reconstruction and the
scanning quality. Concerning the scanning quality, misalignment during the acquisition, noise and ring
artefacts may be corrected or improved by using the appropriate reconstruction parameters.
Prior to the reconstruction procedure, some systems can check the projection images for potential
movements during the scanning procedure. The alignment of the projection images may x these
movements. If sample movements cannot be corrected through the reconstruction procedure, the
scanning process should be repeated.
Following the alignment of the projection images, the reconstruction software creates a histogram with
the frequency the grayscale values representing the density distribution of the micro-CT dataset. The
specic range of the histogram values can display different parts of the organisms (Fig. 11). Generally,
the peaks in the histogram values represent different structures concerning the different densities. Dense
structures such as bones and shells are represented by higher grayscale values compared to soft tissues.
The user can choose to restrict the range of values, to suppress or include specic densities, in order to
emphasize different structures in the reconstruction of the sample. The range of the grayscale values can
also be taken into consideration in order to avoid noise and to isolate unwanted structures (e.g., sample
holder).
During the reconstruction procedure, the reconstruction can be restricted to a specic area of the specimen
by creating a region of interest (ROI). For example, the reconstruction of the jaws of a polychaete can be
achieved by the creation of a ROI which comprises only these target structures (Fig. 12). This method
Specimen Voltage
(kV)
Current
(μΑ) Filter System Source Reference
Fossils in amber 60 unknown none from University of Ghent unknown Dierick et al. 2007
Fossils in amber 120 unknown 1 mm
aluminium from University of Ghent unknown Kehlmaier et al.
2014
Xandarella spectaculum
(Arthropoda) in slab 90 120 0.3 mm Cu Phoenix GE Nanotom tungsten Liu et al. 2015
Lignied plant mesofossils 80 124 0.5 mm
aluminium SkyScan 1172 tungsten Kvaček et al.
2016
Charcoalied plant
mesofossils 40 250 none SkyScan 1172 tungsten Hermanová et al.
2017
Silicied cones 180–200 180–200 unknown GE Phoenix v|tome|x s tungsten Gee 2013
Ediacara fossil Pteridinium
simplex 200 240 none
ultra-high resolution
subsystem of the ACTIS
scanner unknown Meyer et al. 2014
Trace fossil Trichichnus 130 61 1 mm
aluminium SkyScan 1173 tungsten Kędzierski et al.
2015
Shark tooth
(Sphaenacanthus
hybodoides)
180 138 unknown Nikon Metrology HMX-
ST 225 tungsten Abel et al. 2012
Table 6. Examples of scanning parameters for palaeontological samples.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
19
can minimise the time of the reconstruction procedure and the size of the reconstructed dataset. The
reconstruction duration also depends on the computational resources and capacities, the size of the
dataset and the algorithm used (Sutton et al. 2014).
Fig. 11. Histogram of the grayscale value frequency of the scanned specimen (bivalve Musculus
costulatus (Risso, 1826). Each peak represents a different structure (in terms of density) of the scanned
bivalve. Bright grayscale values (representing low densities) are located at the left side of the histogram,
darker values (representing high values) at the right side of the histogram. A. The selection of a range
including all peaks, reveals the more detailed morphology of the bivalve (both soft/low density and
hard/high density structures). B. A restricted range of histogram values removes structures with brighter
values (= low densities). C. A restricted range of histogram values including only one peak reveals only
the darkest values (= most dense structures) of the bivalve which correspond to the shell. Image by
HCMR micro-CT lab.
Fig. 12. Cross-section image without (A) and with (B) a selection of a region of interest (red square) for
the reconstruction of polychaete jaws. Image by HCMR micro-CT lab.
European Journal of Taxonomy 522: 1–55 (2019)
20
Reconstructed data should ideally be saved without any compression or down-sampling (i.e., as 16 bit
TIFF les). However, these les are large, so if storage space is an issue or data are to be shared, the
creation of compressed image formats (e.g., 8 bit PNG, JPG) can be considered – but always taking into
account the detail of information required for the planned analyses. The best lossless image format is
TIFF as it can also store metadata (e.g., voxel size, specimen info, scan parameters); however, different
systems offer different options.
3.5. Visualisation and post-processing
The reconstructed images can be visualized in 3D using volume rendering software. A variety of
products are available (see Table 7). The creation of interactive 3D volumes allows the users to explore
the dataset from any direction and to manipulate its appearance by changing the rendering parameters
(Ruthensteiner et al. 2010). The 3D visualisation of specimens may be used for taxonomic purposes, as
specic structures can be visualised in their original orientation and shape (see Faulwetter et al. 2013a).
A volume can be visualised through volume rendering or through extracting an isosurface (Sutton et al.
2014). Details related to these visualisation methods are presented below.
Table 7 (continued on next page). 3D Volume Rendering software (modied table from Walter et al.
2010 and Abel et al. 2012).
Software Licence
Type URL
Amira commercial www.amira.com
Arivis
(web-based
software)
commercial http://vision.arivis.com/
BioImageXD free http://www.bioimagexd.net
Blender free www.blender.org
Brain Maps
(web-based
software)
free http://brainmaps.org
CTVox free https://www.bruker.com/products/microtomography/micro-ct-software/3dsuite.html
Dragony
free licences
available for
researchers
with non-
commercial
activities/
commercial
http://www.theobjects.com/dragony/
DRISHTI free http://sf.anu.edu.au/Vizlab/drishti/
Fiji (Is Just
ImageJ) free http://ji.sc/
Huygens commercial http://www.svi.nl
ImageJ free https://imagej.nih.gov/ij/
Image-Pro commercial http://www.mediacy.com
Imaris commercial http://www.bitplane.com/
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
21
The post-processing of micro-CT datasets can include simple analyses (e.g., density estimation through
the calculation of the grayscale values, porosity, thickness) or advanced morphometric analysis (e.g.,
geometric morphometrics). The latter requires segmentation of the image (isolation of features of interest
and creation of a geometric surface model – see below).
3.5.1. Volume Rendering
The volume rendering procedure assigns colour and opacity to each voxel according to the grayscale
values of the sample (Kniss et al. 2002). Some ranges of the histogram can be set to be transparent, mostly
to exclude the voxels of the surrounding medium or container. Whenever a structure can be well-dened
by its density/grey level on an histogram, it is easily isolated on a volume rendering image by setting
everything else transparent (Fig. 13). Advanced rendering parameters give the user the opportunity to
apply articial colours and brightness in order to create a realistic/useful rendering.
Fig. 13. Volume rendering of a specimen where the gray level coding for (A) air and (B) air+soft tissues
are transparent. Histograms of the grayscale values are included for both images where the selected
threshold is indicated by the blue line and the opacity curve is indicated by the red line. Image by MNHN.
Software Licence
Type URL
Mimics commercial www.materialise.com/mimics
Octopus commercial https://octopusimaging.eu/
Open
Inventor commercial http://www.openinventor.com/
Simpleware commercial www.simpleware.com
Slice:Drop
(web-based
software)
free http://slicedrop.com/
SPIERS free https://spiers-software.org/
tomviz free http://www.tomviz.org/
VG Studio
Max commercial www.volumegraphics.com
Volocity commercial http://www.improvision.com
VolViewer free http://cmpdartsvr3.cmp.uea.ac.uk/wiki/BanghamLab/index.php/VolViewer
VTK free http://www.vtk.org/
Table 7 (continued).
European Journal of Taxonomy 522: 1–55 (2019)
22
3.5.2. Isosurface rendering
An isosurface is a geometric mesh connecting 3D points of a constant intensity within a volume
(Fig. 14). The thresholding procedure (or binarisation) of the slices, where the grayscale images are
transformed to black and white images, is important for the creation of the 3D model while all voxels are
connected above the thresholding value (for software options see Table 8). The 3D model construction
is based on the marching cubes algorithm (Lorensen & Cline 1987). The calculation might be time-
consuming, depending on the size of the data and the computer capacities. The resulting triangle-mesh
can be visualised and it is suitable for analysis (e.g., shape analysis, volume, geometric morphometrics,
or nite-element modelling). For software packages related to 3D analysis see Table 9.
Software Licence Type URL
Amira commercial www.amira.com
BioImageXD free http://www.bioimagexd.net
Dragony free licences available
for researchers with non-
commercial activities/
Commercial
http://www.theobjects.com/dragony/
DRISHTI free http://sf.anu.edu.au/Vizlab/drishti/
ilasti free http://ilastik.org/
Mimics commercial www.materialise.com/mimics
Octopus commercial https://octopusimaging.eu/
Simpleware commercial www.simpleware.com
SPIERS free https://spiers-software.org/
VG Studio Max commercial www.volumegraphics.com
Table 8. Segmentation software (modied table from Walter et al. 2010 and Abel et al. 2012).
Fig. 14. A marine worm (Polychaeta, Phyllodocidae, Phyllodoce). A. Photograph (CC-BY-SA Hans
Hillewaert). B. Volume rendering. C. Isosurface rendering. Images B and C by the HCMR micro-CT lab.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
23
3.5.3. Segmentation
The main drawbacks of volume rendering and isosurface rendering are that neighbouring structures
cannot be discerned if their densities are too homogeneous and their borders not contrasted enough. In
these cases it is difcult to discern a specic structure among voxels of similar grey value (for example:
visualising specic organs within the body where all organs are of similar densities). Segmentation is
the process of selecting (‘labelling’) voxels of interest in order to visualise them separately of the whole
dataset (or to generate a 3D model of these user-dened structures) (Fig. 15). The segmentation procedure
Table 9. 2D/3D analysis software (modied table from Walter et al. 2010 and Abel et al. 2012).
Software Licence Type URL
Amira commercial www.amira.com
Arivis (web-based software) commercial http://vision.arivis.com/
BioImageXD free http://www.bioimagexd.net
Fiji (Is Just ImageJ) free http://ji.sc/
GOM Inspect free http://www.gom.com/nl/3d-software/gom-inspect.html
Huygens commercial http://www.svi.nl
ImageJ free https://imagej.nih.gov/ij/
Image-Pro commercial http://www.mediacy.com
Imaris commercial http://www.bitplane.com/
MeshLab free http://www.meshlab.net/
MorphoJ free http://www.ywings.org.uk/morphoj_page.htm
Simpleware commercial www.simpleware.com
Stratovan Checkpoint commercial www.stratovan.com
tomviz free http://www.tomviz.org/
VG Studio Max commercial www.volumegraphics.com
Volocity commercial http://www.improvision.com
VTK free http://www.vtk.org/
Fig. 15. Polychaete specimen (Lumbrineris latreillii Audouin & Milne Edwards, 1834) in a composite
rendering showing the location of organs of interest within the animal. Soft tissues are volume-rendered,
jaws were segmented individually and surface-rendered in different colours. The coloured arrows at the
upper left corner indicate the orientation of the scanned specimen in three views (x, y and z axes). Image
by HCMR micro-CT lab, CC-BY Sarah Faulwetter.
European Journal of Taxonomy 522: 1–55 (2019)
24
can be achieved either manually, semi-automatically or automatically – a large number of algorithms
exist for automatic segmentation. Segmentation can be used to include the removal of unwanted objects,
to highlight certain structures with colour or to virtually dissect the sample (Sutton et al. 2014).
3.6. Troubleshooting
Micro-CT scanning is prone to artefacts which can degrade the quality of micro-CT images and the
degree of image distortion can make the micro-CT datasets unusable (Barrett & Keat 2004). Artefacts
can be created by several processes during the acquisition of micro-CT images. The most common
artefacts which are encountered on the micro-CT images are: a) beam hardening artefacts, b) ring
artefacts, c) noise, d) partial volume, e) motion artefacts and f) metal artefacts. The avoidance or the
correction of these artefacts is important to improve the resulting micro-CT image, but sometimes the
effects may be irreversible, so that the scan will have to be repeated.
3.6.1. Beam hardening artefacts
Beam hardening usually occurs when an object consists of different parts with different attenuation
coefcients (especially with high densities). This refers to the fact that the beam which penetrates the
object becomes harder (increased average energy) as the lower energy X-rays are absorbed more rapidly
than the higher-energy X-rays (Barrett & Keat 2004). The result of the beam hardening effect is that the
object appears to be denser (increased brightness) at the edges than the centre (decreased brightness)
creating cupping artefacts (Roche et al. 2010; Abel et al. 2012; Fig. 16). A simple approach to solve this
Fig. 16. Scan of teeth without (A) and with (B) software beam hardening correction. In yellow, cupping
artefacts increase the reconstructed density at the edges (see plots of gray values along the yellow lines)
and decrease it in the centre of the object. In blue, streaking artefacts create dark or white lines between
structures. Images by MNHN.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
25
problem is to use an X-ray beam of higher energy to ensure that beam hardening is negligible, but this
solution is applicable only for small samples (Ketcham & Carlson 2001). Beam hardening effects can
also create streaking artefacts which are shown as dark and light streaks around very dense structures
(Fig. 16). According to Sutton et al. (2014), the reduction of the streaking artefacts can be achieved by
using different scanning media (e.g., water, sand) which act as a lter and smoothen the effect, as well as
through an increase of projection images and a decrease of the exposure time. Beam hardening artefacts
can also be minimised by using a metal lter (e.g., aluminium, copper) (Ketcham & Carlson 2001;
Barrett & Keat 2004; Meganck et al. 2009; Abel et al. 2012; Sutton et al. 2014), which removes the low
energy photons during the scanning procedure. However, this can attenuate the X-ray intensity to some
degree, thus leading to greater image noise unless longer acquisition times are used (Ketcham & Carlson
2001). Concerning cupping artefacts, the use of a combination of aluminium and copper lters is the
most effective way for the reduction of these kind of artefacts (Meganck et al. 2009; Hamba et al. 2012).
Beam hardening effects can be also reduced to some extent by using the beam hardening correction of
the reconstruction software, if provided (Fig. 16).
3.6.2. Ring artefacts
Ring artefacts are light and dark circles (circular artefacts) on the reconstructed images (Fig. 17A) as a
result of calibration deciency of the detector where the X-ray sensitivities of the detector elements are
distinguished (Davis & Elliott 1997; Barrett & Keat 2004; Sutton et al. 2014). In some occasions ring
artefacts are caused by changes in temperature or beam strength, and this can be solved by carefully
controlling experimental conditions or by frequent recalibrations (Ketcham & Carlson 2001; Barrett &
Keat 2004). The reduction of ring artefacts can be also achieved through a at-eld correction (Barrett &
Keat 2004; Sutton et al. 2014), although such artefacts may still persist as a result of the beam hardening
effect and the high-spatial-frequency variations in the thickness of the scintillator detector (Vågberg
et al. 2017). Furthermore, depending on the system, ring artefacts can be reduced by using the selection
of the ‘ring artefacts correction’ during the reconstruction (Fig. 17B) or the selection of the ‘random
movement’ option during the scanning procedure (Sutton et al. 2014).
Fig. 17. Scan of a marine worm (polychaete) without (A) and with (B) ring artefacts correction during
the reconstruction procedure. The red square indicates the presence of ring artefacts which are reduced
in (B) following the ring artefacts correction. Images by HCMR micro-CT lab.
European Journal of Taxonomy 522: 1–55 (2019)
26
3.6.3. Noise
Noise (or quantum noise) is caused due to statistical error of low photon counts and it is presented as
scattered bright and dark streaks in the slices (Fig. 18A) (Boas & Fleischmann 2012). This kind of artefact
can be minimised by using higher beam intensities (increased mAs) which result in better signal-to-noise
ratios (SNR) (Boas & Fleischmann 2012; Sutton et al. 2014). Furthermore, an increased source voltage
(Kachelrieß 2008; Hsieh 2009; Boas & Fleischmann 2012) and an augmented number of projection
images (Sutton et al. 2014) may reduce noise. Low exposure times can result in high noise, so increased
exposure times and the use of lters may reduce these artefacts in dense samples (Sutton et al. 2014).
Furthermore, frame averaging improves the SNR (Sutton et al. 2014). Noise can also be created by
physical limitations of the system (e.g., electronic noise in the detector panel), but the reduction of this
kind of noise by the operators is difcult (Hsieh 2009). Additional noise can be introduced during the
reconstruction procedure. The selection of appropriate reconstruction algorithms and parameters (e.g.,
selection of Gaussian lter) can minimise the noise effect (Fig. 18B) (Hsieh 2009; Sutton et al. 2014).
3.6.4. Partial volume
The partial volume effect creates artefacts through the fact that the average attenuation coefcient within
a voxel represents an average grayscale value (Barrett & Keat 2004; Abel et al. 2012; Sutton et al.
2014). The result of the partial volume effect is shading artefacts in the image (Barrett & Keat 2004).
Sutton et al. (2014) mention that if anatomical structures are close to the voxel size, then the partial
volume effect will be more intense. The use of maximum magnication can decrease partial volume
artefacts, but if this effect still remains after adjusting the magnication the use of a detector with greater
dimensions may be needed (Sutton et al. 2014).
3.6.5. Motion artefacts
Sample motion during the image acquisition creates artefacts which appear as shades or streaks in the
reconstructed images (Fig. 19) (Barrett & Keat 2004). Motion artefacts can be reduced by calculating
and correcting the X/Y pixel shifts during the reconstruction procedure (Salmon et al. 2009). However,
ideally, sample movements should be minimised by stabilising the sample in an appropriate sample
holder (see Section 3.2). Similar artefacts can also occur as a result of sample shrinkage; in this case an
X/Y correction cannot be performed. The use of a liquid as a scanning medium or the use of Paralm®
can prevent sample shrinkage and therefore the resulting artefacts.
Fig. 18. Scan of a bivalve with (A) and without (B) noise after the selection of the appropriate parameters
during the reconstruction procedure. Images by HCMR micro-CT lab.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
27
3.6.6. Metal artefacts
Micro-CT imaging of natural history specimens can be challenging when the specimen contains metal
particles which are visualised in the form of a ‘metal artefact’. These metal particles are common in
many geological samples as natural inclusions or part of the chemical composition of the sample.
However, metal is also used as a support medium when skeleton parts are assembled in their natural
structure. Metal pins are also used to mount entomological samples to mount these in an insect drawer.
The metal artefact is especially observed when there is a large difference between the attenuation of the
metal part and the sample of interest (Figs 20–21). Usually, the artefact is less obvious when scanning
at higher kVp or when the difference in attenuation is less pronounced (e.g., bone and metal) (Fig. 22).
Scanning at high kVp is not possible for soft bodied samples like insects. Metal artefacts may be the
result of several causes like noise, beam hardening, non-linear partial volume effect, and scatter (de
Man 2001). Visually, the metal artefact looks like an overexposed part on the micro-CT image slice with
many streaks and star-like bright lines (Figs 20B, 22B). In some cases reconstruction software contains
a built-in Metal-Deletion-Technique (Boas & Fleischmann 2012) that reduces or completely removes
the metal artefact. Tests with this algorithm in the Xact reconstruction software (RX Solutions) proved
successful for a leopard skull bearing metal screws, but failed for a scan with a pinned beetle.
Fig. 19. Scan of a marine worm (polychaete) with motion artefacts. The structures are not clearly dened
due to specimen movement during the scanning procedure. Image by HCMR micro-CT lab.
European Journal of Taxonomy 522: 1–55 (2019)
28
Fig. 20. A. Part of a pinned Omorgus gigas (Harold, 1872) beetle a few hundred slices away from
the pinned area. B. The normal morphology of the beetle is no longer visible due to the metal artefact
appearing in the pinned area. Image by the Royal Belgian Institute of Natural Sciences (RBINS) / DIGIT-3
Belspo, CC-BY-NC-ND Jonathan Brecko.
Fig. 21. A. 3D model of the Omorgus gigas (Harold, 1872) beetle after a quick segmentation, including
the metal artefact. B. 3D model of the same specimen after manual removal of the pin in the Dragony
software (http://www.theobjects.com/dragony/). Clicking on the image opens the 3D model. Photo
courtesy of the Royal Belgian Institute of Natural Sciences (RBINS) / DIGIT-3 Belspo, CC-BY-NC-ND
Jonathan Brecko.
Fig. 22. A. Monkey vertebra without metal support. B. A metal artefact (yellow arrow) is created due to
the metal rod used to support a series of vertebrae on a mounted skeleton. Photo courtesy of the Royal
Belgian Institute of Natural Sciences (RBINS) / DIGIT-3 Belspo, CC-BY-NC-ND Jonathan Brecko.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
29
4. Use cases
4.1. Zoological samples
The identication of zoological specimens is usually based on external, internal or on the combination of
external and internal morphological characteristics. The 3D nature of the different structures is important
for the classication of the organisms (Boistel et al. 2011; Faulwetter et al. 2013a). The creation of 3D
models at a resolution of a few microns, the accessibility of micro-CT scanners, the low cost and the use
of these datasets for morphometric, functional, ecological and developmental analysis reveal the utility
of this technology for zoological studies (Faulwetter et al. 2013a; Fernández et al. 2014). According to
Boistel et al. (2011), different scans – which are referred to as ‘morphoscans’ – should be included in a
reliable 3D library with reference models (or gold standards) including intra-species variations and key
references established by the community.
Micro-CT is characterised as a non-destructive technology, but potential damage of genetic material due
to the X-ray radiation should be considered. Concerning preserved zoological specimens, studies on the
effect of X-rays on bird skins (Paredes et al. 2012) and on invertebrates (Faulwetter et al. 2013a) revealed
that exposure to X-rays did not cause damage to the specimen DNA, at least not to the investigated gene
(i.e., 16S rRNA). Paredes et al. (2012) speculated that the preservation procedures used on the specimen
may cause more damage to the DNA quality than the X-rays. Similar studies on the effects of the micro-
CT technology on the DNA quality of a variety of organisms in different preservation conditions should
be carried out before this technology can be safely used on valuable museum specimens and especially
on type material (Paredes et al. 2012; Faulwetter et al. 2013a).
Micro-CT technology has been used in morphological and anatomical studies, such as the creation of a
3D interactive model of the jaw musculature of the American alligator, from which several researchers
who are interested in feeding ecology and evolutionary morphology have proted (Holliday et al. 2013).
Embryonic imaging as well as the quantitative analysis of the organs and tissues of chick embryos
has been achieved through micro-CT (Kim et al. 2011), revealing the benets of this technique for
developmental biology (Metscher 2009b; Kim et al. 2011).
Another example of the use of micro-CT in zoological specimens is the assessment of the effects
of ocean acidication on calcied structures of marine invertebrates, revealing the utility of this
technique in ecological studies (Keklikoglou et al. 2015; Chatzinikolaou et al. 2017). Furthermore,
Faulwetter et al. (2013a) have shown advantages of the technique for taxonomic studies through the 3D
visualisation of polychaetes. The latter authors demonstrated the importance of the micro-CT technique
in the creation of ‘cybertypes’, as a potential addition to the current type material of the collections,
with implications in our traditional ways of carrying out systematics and for the International Code of
Zoological Nomenclature (Godfray 2007).
4.2. Botanical samples
Micro-computed tomography can be still considered as an under-used technique in botanical research
(Staedler et al. 2013). Botanical samples often contain soft tissues such as leaves, owers or fruit tissues
with a low X-ray attenuation, and imaging of such parts may require contrast enhancement to increase
the quality of images. Best results are usually obtained from samples that consist of different tissue types
with different densities, such as seeds, fruits, woods, etc, as these allow a better discrimination between
individual organs (Leroux et al. 2009). Usually, secondary cell walls are more easily detected than soft
tissues (Leroux et al. 2009).
The technique is very suitable for visualising delicate anatomical structures such as embryos, meristems,
owers, and even individual cell walls (Dhondt et al. 2010; van der Niet et al. 2010; Johnson et al. 2011).
European Journal of Taxonomy 522: 1–55 (2019)
30
The applications for micro-CT in botanical research are very diverse. Among the earliest applications
of micro-CT in botany were root studies (e.g., Crestana et al. 1986), and the literature in this eld is
abundant (see review in Mooney et al. 2012). Roots, but also whole plants, can in many cases also
be scanned in vivo. This has allowed researchers to perform growth and repair experiments in living
studies, and even record time-lapse 3D images (e.g., Brodersen et al. 2010, 2013; Dhondt et al. 2010;
Paquit et al. 2011; Knipfer et al. 2015). However, the ionising effect of X-rays has to be taken into
account, which may cause growth inhibition or untypical developments (e.g., Dhondt et al. 2010), so in
vivo studies need to be planned carefully.
Micro-CT has furthermore been used for morphometric analyses of plants, which form the background
for e.g., comparative morphology (van der Niet et al. 2010; Wang et al. 2015), food production (Li
et al. 2011) or to address evolutionary questions (Miller & Venable 2003). Due to the high resolution of
modern micro-CT scanners – and the large and well-dened cell structures in many plant tissues – cellular
structures can often be discerned in plant scans. These ne structures aid in understanding physiological
and functional processes in plants (Gamisch et al. 2013; Pajor et al. 2013; McElrone et al. 2013), as
cellular architecture profoundly affects the physiology of plant tissues (Pajor et al. 2013). Finally, apart
from studying soft plant tissues, micro-CT has been extensively used to study the density and porosity
of woods, including plant hydraulics and xylem studies (e.g., Brodersen et al. 2011; Cochard et al. 2015;
see also Brodersen & Roddy 2016 for an overview).
4.3. Palaeontological samples
Morphological information on fragile, rare and valuable palaeontological specimens has become
accessible through micro-CT imaging (Abel et al. 2012). Palaeontological specimens are usually
embedded in sediment or in amber. The investigation of these inclusions using scanning electron
microscopy (SEM) or transmission electron microscopy (TEM) provides high quality images, but these
methods require the destruction of the material (Greco et al. 2011; Görög et al. 2012). The internal
structure of palaeontological specimens is frequently important for their taxonomic classication (Görög
et al. 2012). For palaeontological objects, such as pyritised fossils which are difcult to conserve, micro-
CT seems to be the only solution for the visualisation of their internal morphology, as this technique does
not require the removal of the conservation matrix (DeVore et al. 2006). Palaeontological specimens in
sedimentary rock or amber matrix can thus be ‘virtually’ extracted using micro-CT (Abel et al. 2012;
Konietsko-Meier & Schmitt 2013).
The choice of micro-CT for visualising palaeontological specimens should take into account the required
energy that is needed to penetrate such dense samples as well as the required contrast between specimen
and matrix (Sutton et al. 2014). Furthermore, the size of the fossils and the size of the structures to be
visualised might be an important constraint for the choice of this technology. Micro-CT is suitable for
the majority of fossils whose size ranges from 2 mm to 200 mm (depending on the micro-CT system)
(Sutton et al. 2014). According to Rahman & Smith (2014), micro-CT is suitable for fossil groups such
as vertebrates, invertebrates, plants, microfossils, trace fossils and for a variety of their preservation
types (including altered preservation, cast, mold, original and permineralisation).
The use of micro-CT in palaeontological collections is mostly related to the 3D visualisation and 3D
analysis of specimens for taxonomic (e.g., Penney et al. 2007; Briguglio & Benedetti 2012), evolutionary
(e.g., Koenigswald et al. 2011; Garwood et al. 2014), histological (e.g., Konietsko-Meier & Schmitt
2013) and palaeoecological purposes (e.g., Wan et al. 2014; Yang et al. 2015). Micro-CT has been
used to study the fossil inclusions of amber and this technology has revealed features and internal
structures which are important for the taxonomical classication of such specimens (Dierick et al.
2007; Kvaček et al. 2018a). Furthermore, an example of the utility of micro-CT in morphological and
paleobiological studies is the study of the skeleton of an Eocene–Oligocene primate which allowed
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
31
the visualisation of features related to the life history, locomotion and diet (Franzen et al. 2009). The
digital restoration of the 3D morphology of the Ediacara fossil Pteridinium simplex Gürich, 1930 using
micro-CT revealed information concerning its ecology and its phylogenetic afnity (Meyer et al. 2014).
Micro-CT has successfully been used for studies of plant and insect mesofossils (Kvaček et al. 2016;
Heřmanová et al. 2017; Kvaček & Heřmanová 2017; Kvaček et al. 2018a, 2018b). Application of the
method for lignied fossil plants embedded in sediment is not always easy. However, if the fossil is
not completely compressed and contrast between the fossil and the sediment is high enough, it can be
successful (Kvaček & Heřmanová 2017; Kvaček et al. 2018b). Charcoalied plant fossils consist of
charred organic matter and air, providing excellent contrast between the air lling internal parts of cells
and the charcoalied cell walls. The use of micro-CT to study the anatomy of fruits, seeds and insect
eggs has been proved as a fast and user friendly tool for investigating their internal structures, which are
crucial for their interpretation (Kvaček et al. 2016; Kvaček & Heřmanová 2017).
4.4. Geological (mineral) samples
Geological applications of X-ray computed tomography include interior examination of one-of-a-kind
fossils or meteorites, textural analysis of igneous and metamorphic rocks, geometric description and
quantication of porosity and permeability in rocks and soils, and any other application demanding
three-dimensional data that formerly required physical serial sectioning (Ketcham & Carlson 2001).
Micro-CT scans of rocks are virtual models with grayscale values representing primarily the X-ray
attenuation of different minerals or other features, such as pores, in the rock (Wellington & Vinegar
1987). This method offers the means to study the internal materials and geometrics of rare or valuable
specimens, such as meteorites, fossils or archaeological artefacts, which should not be destructively
sectioned. Even in cases where sectioning is allowed, the use of micro-CT can eliminate the time spent
and the laboriousness traditionally faced during the serial sectioning of geological specimens. The
digital character of a micro-CT dataset allows interactive manipulation of the data, better visualisation
and animation and easier measurement of dimensions (Carlson et al. 2003; Abel et al. 2011).
Micro-CT is a useful tool for the visualization of fractures, pores or layers which are characterised by
different densities (Cnudde et al. 2006). According to Stanley (1992), among the types of rocks that can
be examined using micro-CT are:
– igneous rocks formed by cooling and hardening of molten material (magma), which are composed
by interlocking grains, each consisting of a particular mineral).
– rocks formed from sediments (mineral grains) that are deposited at the earth’s surface by water, ice
or air.
– metamorphic rocks formed by the alteration of rocks within the earth under conditions of high
temperature and pressure, which are characterised by minerals and textures arrayed in parallel wavy
layers.
– crystalline rocks which are igneous and metamorphic rocks formed at high temperatures.
High-resolution X-ray computed tomography has been used to reveal the sizes and the three-dimensional
spatial disposition of porphyroblasts in metamorphic rocks, which can indicate the atomic-scale
processes that control crystal nucleation and growth (Denison et al. 1997). Petroleum engineers have
used CT data to study two-uid coreood experiments in reservoir lithologies (Wellington & Vinegar
1987). Meteorite investigations have progressed from solely nding inclusions to mapping out the shape
and size distributions of their mineralogical components, providing textural clues about their origins
(Kuebler et al. 1999). The 3D images produced using micro-CT technology allow the study of the real
3D petrography of meteorites and the visualization of a larger percentage of the meteorite (Hezel et al.
2013).
Another example of the use of micro-CT in petrology studies is the analysis of temporal, geographical
and species-specic variations in aked stone tool morphology, which attempts to explain the evolution
European Journal of Taxonomy 522: 1–55 (2019)
32
of cognition, culture and human behaviour. Abel et al. (2011) studied aked stone tools used by early
humans from at least 2.6 million years ago. Using micro-CT, they were able to visualise the key
features of percussion, which distinguish akes intentionally made by humans (artefact) from the ones
created naturally (geofact). Also, they could recreate missing akes from retted groups of material by
visualising void spaces (Abel et al. 2011).
5. Data curation
Given the importance of virtual specimens for future research – especially if datasets are designated
as cybertypes for species descriptions (see e.g., Akkari et al. 2015) – data management (or curation)
of micro-CT datasets should be performed with the same care as curating physical specimens.
Currently, however, there are no universally accepted guidelines or standards concerning the curation,
documentation and dissemination of micro-CT datasets. A recent overview and set of recommendations
has been provided by Davies et al. (2017). A few general suggestions are listed below, which should
be taken into consideration by micro-CT laboratories wanting to develop or improve data management
practices.
Generally, data management / curation activities can be classied into three broad categories:
a) documentation; b) data organization, storage and archival practices and c) data dissemination.
5.1. Documentation
The documentation of the dataset through metadata is crucial. Metadata place the dataset into context,
make it discoverable and retrievable, provide information on provenance, terms for re-use and document
the steps performed to create the dataset. The required level of detail of the metadata is determined by a
variety of factors, such as internal guidelines of the institution, every-day practices of the laboratory and
demands of users, trying to strike a balance between the need for a full documentation of the dataset with
a high level of detail, and the effort, time and costs required to create this documentation.
Currently, no metadata standards or formats exist specically for natural history specimens imaged
through micro-CT. Several existing standards could be adapted to adequately describe these data, such
as the DICOM (Digital Imaging and Communications in Medicine; http://dicom.nema.org), ISA-TAB
(Investigation–Study–Assay; http://isa-tools.org/; Sansone et al. 2012) or HDF5 (Hierarchical Data
Format 5; https://www.hdfgroup.org/) format. DICOM is a le format and data exchange protocol
extensively used in medical imaging and contains the metadata embedded together with the images.
While the metadata can be adapted to individual needs, and thus could be tailored towards 3D natural
history specimen data, DICOM is not straightforward to learn, dedicated software is required and
DICOM is not produced directly by many scanner models. The ISA-TAB standard provides exible
options to describe protocols and parameters and to combine metadata terms. It has at least once been
applied to document micro-CT data (Stoev et al. 2013), although in a simplied form. The Hierarchical
Data Format (HDF5) is a generic format suitable for large datasets comprised of heterogeneous subsets,
including user-dened embedded metadata. The HDF5 format has been used for tomographic data (e.g.,
De Carlo et al. 2018; Mancini et al. 2018), but its use requires some expertise and dedicated software
is needed.
However, many laboratories may decide to store metadata in a custom-made database or even in text
les that are distributed along with the data. While a central database which includes documentation for
all datasets allows for efcient search and retrieval of the datasets and enhances discoverability, care
needs to be taken when data are re-distributed. In this case the metadata need to be extracted from the
database and included together with the image les (e.g., as a text le, or embedded in, e.g., DICOM
or HDF5 les). In the case of custom-made documentation systems, metadata should at least include:
– unique identier for the dataset. Ideally, this should be a globally unique identier, but in case the
datasets are used only internally, they can be unique within the lab / institution
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
33
– description of specimen, including a link to an identier for the specimen, e.g., a catalogue number
– purpose of dataset
– person(s) involved (for each action, if this level of detail is required)
– preparations of specimen (e.g., staining) which inuence the results
– date/time of scanning and all scanning parameters (often recorded automatically by the machine)
– description of / Link to the storage location of the dataset(s)
– usage restrictions / licenses / copyright
– information on provenance / changes / version of the dataset (or a “last modied” date)
– links to or identiers of derived publications or datasets
5.2. Data organisation, storage and archival practice
While careful metadata annotation can help to organise and retrieve the actual datasets, management of
the actual les is equally important. Tomography data pose certain challenges due to their large volume,
requiring adequate long-term storage, reliable backup systems and mechanisms to organise and retrieve
les on storage media.
Depending on the responsibilities of the laboratory, different data products may have to be managed: In
some laboratories, only the raw data (projection images) are kept, and everything from the reconstructed
cross sections to images, analyses and other derivative data falls under the responsibility of the customer/
scientist; other practices include keeping also the reconstructed set of images and/or derivative data. A
le organisation and storage system should take these responsibilities into account by organising data
in a consistent directory structure with consistent le names which can be intuitively and comfortably
understood and used (e.g., Yakami et al. 2011; Faulwetter et al. 2015).
Disk storage space and long-time archival practice has to be carefully accounted for. Several options for
storage exist, but all of them have advantages and disadvantages:
– external and portable hard disks are easy to use but not very reliable. If hard disks are used, those
supporting RAID (Redundant Array of Independent Disks) technology can provide extra protection
against data loss.
– Network Attached Storage (NAS) servers are currently the best option. They can provide scalability
and fault tolerance beyond that of discrete hard disks.
– tape library storage is still a common backup option, but recovery of data can be time-consuming and
difcult.
– off-site storage (i.e., in the cloud) should be considered as protection against local disasters, but the
large datasets (often several terabyte of data) result in long transmission times over the internet.
Whether online storage is a feasible option will therefore depend on the practices of the laboratory
(e.g., frequency and volume of data produced; type of data to be archived).
Laboratories with limited storage capabilities may also consider using data compression methods (see,
e.g., Mancini et al. 2018 for an overview).
5.3. Data dissemination and publication
5.3.1. Data publication
The publication of research data, either to underpin scientic publications or as a general good practice
of data sharing, has gained increasing acceptance over the last years. Sharing tomography data, however,
can be still a challenge due to the large data volume. Data can be shared either through an institutional
le sharing server or be published through a public repository. Available repositories differ in their
scope and features. The choice depends on the type and size of the data to be shared, on whether journal
policies apply (e.g., when data accompany a research article) and potentially on institutional practices.
An overview of repositories is given in Table 10.
European Journal of Taxonomy 522: 1–55 (2019)
34
Table 10. Overview of repositories for 3D data.
Data repository Scope / Restrictions URL
DigitalMorphology primarily hosts data from the Digimorph group
and collaborators; for contributions the project
administrators should be contacted. Available data
are QuickTime animations of CT stacks, movies of
3D volume rendering of specimens and STL surface
models.
http://digimorph.org
Dryad generic data repository. Not tailored towards large
les, although they are accepted. Charges a publishing
fee (currently $120 per dataset), unless waivers apply.
All data are released under a CC-Zero waiver (public
domain, copyright removed). Assigns DOIs to the
data.
http://www.datadryad.org
Figshare generic data repository. Supports les up to 5 GB free
(beyond that, institutional services can be purchased).
License can be chosen by user. Assigns DOIs to the
data.
https://gshare.com
GigaDB accepts almost only data accompanying articles in
GigaScience. Data can, however, be described as a
DataNote and then included in the repository. Focuses
on large-size data, i.e., volumetric data. Encourages
publication of projection and reconstructed data. All
data are released under a CC-Zero waiver (public
domain, copyright removed). Assigns DOIs to the
data.
http://gigadb.org/
LifewatchGreece
micro-CTvlab
accepts volumetric micro-CT datasets, provides
extensive metadata and on-the-y viewer of
volumetric datasets. Several features, e.g., raw data
download, still under construction. Does not assign
DOIs.
https://microct.portal.lifewatchgreece.eu
MorphDBase media database which also develops ontologies that
describe morphological terminology. Not specically
targeted at 3D data but accepts media les in general.
https://www.morphdbase.de
Morphomuseum publishes data accompanying journal articles in
MorphoMuseum, but not only. Focuses on surface
models, but volume data can also be published if they
accompany surface data. Focus on vertebrates, but
does not explicitly exclude other taxa. Assigns DOIs.
Only license available is CC-BY-NC.
https://morphomuseum.com/
Morphosource targeted at 3D biodiversity data (volumetric and
surface). File formats include tiff, dicom, stanford ply,
and stl. Does not assign DOIs automatically. Choice
of license.
http://morphosource.org
Phenome10K provides free 3-D image data of biological and
paleontological specimens to the academic
and educational (non-commercial) community.
Supports all types of 3-D images, including surface
scans, CT-scans and MRIs accompanied by their
metadata, and images can be downloaded as STL les.
Does not assign DOIs. Only license available is CC-
BY-NC.
http://phenome10k.org/
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
35
Both volumetric datasets (reconstructed images and/or projection images) as well as surface models
can be shared and are of interest to the scientic, cultural/artistic and education/science communication
community. When sharing data, a suitable licence should be chosen that allow others to re-use and
process the data. Creative Commons licenses (http://www.creativecommons.org), ideally CC-BY
(attribution required) or even CC-Zero (release into the public domain) are the most suitable licenses for
sharing scientic data (Hagedorn et al. 2011).
5.3.2. Tools for outreach and interaction with 3D data
Apart from making 3D datasets available for download, tools exists for remote on-the-y interaction
of users with the data. These tools are useful for giving users an overview of the data without having to
download them rst, and/or provide a suitable tool for outreach and education.
Surface models can easily be rendered in a browser (e.g., 3D Hop, http://vcg.isti.cnr.it/3dhop/) and allow the
user to interact (rotate, zoom, change colours or lights) with them, and multiple technological solutions and
ready-made portals exist (e.g., WebGL, Java, Flash). Rendering volumetric data online is still a challenge,
and solutions are limited. Arivis Web View (https://www.arivis.com/de/imaging-science/arivis-webview)
is a commercial solution and provides online 2D and 3D rendering even via limited bandwidth. 3D slicer
(https://www.slicer.org/) is free software which likewise allows online rendering of volume data, with a
variety of features to control opacity and colouring.
In addition to browser-based tools, several applications for mobile phones exist as well (e.g., the free CTVox
app, https://www.bruker.com/products/microtomography/micro-ct-software/3dsuite.html; ImageVis3D,
http://www.sci.utah.edu/cibc-software/imagevis3d.html; Volumize, http://www.volumize.be/; or
DroidRender https://www.facebook.com/droidrender). These applications do, however, not work on the
full high-resolution dataset, but on a downscaled version, as neither memory nor graphics cards on
mobile devices are as powerful as those on dedicated desktop computers.
Another tool for interaction with 3D data, and especially suited for exhibitions, are touch-screen
based volume rendering software packages which provide a simplied user interface, allowing
children and non-technical users to explore 3D datasets. A commercial solution is the InsideExplorer
(http://www.interspectral.com/inside-explorer/), which also provides support in setting up exhibitions.
Similar features, but without technical support, are provided by the DrishtiPrayog software
(http://nci.org.au/systems-services/scientic-visualisation/visualisation-services/drishti-prayog/).
Acknowledgements
The creation of this handbook was funded by the EU FP7 programme SYNTHESYS3 (FP7 -312253)
and further supported by the LifeWatchGreece infrastucture (ESFRI - 384676) and BIOIMAGING-
GR (ESFRI - 5002755). The authors would like to thank all the SYNTHESYS participants for their
contribution to this handbook and especially Vincent S. Smith (Natural History Museum, UK) and
Elspeth M. Haston (The Royal Botanic Garden Edinburgh, UK) for their support, and Alex Ball and
Farah Ahmed (Natural History Museum, UK) for their suggestions. The authors would also like to thank
Thaddaeus Buser and one anonymous reviewer for providing comments and suggestions that improved
the manuscript.
References
Abel R.L., Partt S., Ashton N., Lewis S.G., Scott B. & Stringer C. 2011. Digital preservation and
dissemination of ancient lithic technology with modern micro-CT. Computers & Graphics 35 (4): 878–
884. https://doi.org/10.1016/j.cag.2011.03.001
European Journal of Taxonomy 522: 1–55 (2019)
36
Abel R.L., Laurini C.R. & Richter M. 2012. A palaeobiologist’s guide to ‘virtual’ micro-CT preparation.
Palaeontologia Electronica 15 (2): 6T.
Akkari N., Enghoff H. & Metscher B.D. 2015. A new dimension in documenting new species: high-
detail imaging for myriapod taxonomy and rst 3D cybertype of a new millipede species (Diplopoda,
Julida, Julidae). PLoS ONE 10 (8): e0135243. https://doi.org/10.1371/journal.pone.0135243
Alba-Tercedor J. 2012. Studying the anatomy of wet specimens of mayies of the genus Baetis (Insecta:
Ephemeroptera) by scanning them into a liquid with the Skyscan 1172 high resolution micro-CT.
SkyScan microCT 2012 User Meeting, Belgium: 188–191. Skyscan, Brussels.
Alba-Tercedor J. & Sánchez-Tocino L. 2011. The use of the SkyScan 1172 high-resolution micro-CT to
elucidate if the spicules of the “sea slugs” (Mollusca: Nudibranchia, Opisthobranchia) have a structural
or a defensive function. SkyScan microCT 2011 User Meeting, Belgium: 113–121. Skyscan, Leuven.
Barrett J.F. & Keat N. 2004. Artifacts in CT: recognition and avoidance 1. Radiographics 24 (6): 1679–
1691. https://doi.org/10.1148/rg.246045065
Betz O., Wegst U., Weide D., Heethoff M., Helfen L., Lee W.K. & Cloetens P. 2007. Imaging applications
of synchrotron X-ray phase-contrast microtomography in biological morphology and biomaterials
science. I. General aspects of the technique and its advantages in the analysis of millimetre-sized arthropod
structure. Journal of Microscopy 227 (1): 51–71. https://doi.org/10.1111/j.1365-2818.2007.01785.x
Blagoderov V., Kitching I.J., Livermore L., Simonsen T.J. & Smith V.S. 2012. No specimen
left behind: industrial scale digitization of natural history collections. ZooKeys 209: 133–146.
https://doi.org/10.3897/zookeys.209.3178
Boas F.E. & Fleischmann D. 2012. CT artifacts: causes and reduction techniques. Imaging in Medicine
4 (2): 229–240.
Boistel R., Swoger J., Kržič U., Fernandez V., Gillet B. & Reynaud E.G. 2011. The future of
three-dimensional microscopic imaging in marine biology. Marine Ecology 32 (4): 438–452.
https://doi.org/10.1111/j.1439-0485.2011.00442.x
Bribiesca-Contreras F. & Sellers W.I. 2017. Three-dimensional visualisation of the internal anatomy
of the sparrowhawk (Accipiter nisus) forelimb using contrast-enhanced micro-computed tomography.
PeerJ 5: e3039. https://doi.org/10.7717/peerj.3039
Briguglio A. & Benedetti A. 2012. X-ray microtomography as a tool to present and discuss new taxa:
the example of Risananeiza sp. from the late Chattian of Porto Badisco. Rendiconti Online Società
Geologica Italiana 21: 1072–1074.
Brodersen C.R. & Roddy A.B. 2016. New frontiers in the three-dimensional visualization of plant
structure and function. American Journal of Botany 103: 1–5. https://doi.org/10.3732/ajb.1500532
Brodersen C.R., McElrone A.J., Choat B., Matthews M.A. & Shackel K.A. 2010. The dynamics of
embolism repair in xylem: In vivo visualizations using high-resolution computed tomography. Plant
Physiology 154: 1088–1095. https://doi.org/10.1104/pp.110.162396
Brodersen C.R., Lee E.F., Choat B., Jansen S., Phillips R.J., Shackel K.A., McElrone A.J. &
Matthews M.A. 2011. Automated analysis of three-dimensional xylem networks using high-resolution
computed tomography. The New Phytologist 191: 1168–1179.
https://doi.org/10.1111/j.1469-8137.2011.03754.x
Brodersen C.R., McElrone A.J., Choat B., Lee E.F., Shackel K.A. & Matthews M.A. 2013. In vivo
visualizations of drought-induced embolism spread in Vitis vinifera. Plant Physiology 161: 1820–1829.
https://doi.org/10.1104/pp.112.212712
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
37
Burrow C.J., Jones A.S. & Young G.C. 2005. X-ray microtomography of 410 million-year-old optic
capsules from placoderm shes. Micron 36 (6): 551–557. https://doi.org/10.1016/j.micron.2005.05.005
Carlson W.D. & Denison C. 1992. Mechanisms of porphyroblast crystallization: results from high-
resolution computed X-ray tomography. Science 257 (5074): 1236–1239.
https://doi.org/10.1126/science.257.5074.1236
Carlson W.D., Rowe T., Ketcham R.A. & Colbert M.W. 2003. Applications of high-resolution X-ray
computed tomography in petrology, meteoritics and palaeontology. In: Mees F., Swennen R., Van
Geet M. & Jacobs P. (eds) Applications of V-ray Computed Tomography in the Geosciences 215 (1):
7–22. Geological Society Special Publications, London. https://doi.org/10.1144/GSL.SP.2003.215.01.02
Chatzinikolaou E., Grigoriou P., Keklikoglou K., Faulwetter S. & Papageorgiou N. 2017. The
combined effects of ocean acidication and elevated temperature on the shell density of two
gastropod species using micro-CT imaging. ICES Journal of Marine Science 74 (4): 1135–1149.
https://doi.org/10.1093/icesjms/fsw219
Cnudde V., Masschaele B., Dierick M., Vlassenbroeck J., Van Hoorebeke L. & Jacobs P. 2006.
Recent progress in X-ray CT as a geosciences tool. Applied Geochemistry 21 (5): 826–832.
https://doi.org/10.1016/j.apgeochem.2006.02.010
Cochard H., Delzon S. & Badel E. 2015. X-ray microtomography (micro-CT): A reference technology
for high-resolution quantication of xylem embolism in trees. Plant, Cell and Environment 38 (1):
201–206. https://doi.org/10.1111/pce.12391
Crestana S., Cesaero R. & Mascarenhas S. 1986. Using a computer assisted tomography miniscanner in
soil science. Soil Science 142: 56–61.
Davies T.G., Rahman I.A., Lautenschlager S., Cunningham J.A., Asher R.J., Barrett P.M., Bates K.T.,
Bengtson S., Benson R.B., Boyer D.M., Braga J., Bright J.A., Claessens L.P.A.M., Cox P.G., Dong X.-P.,
Evans A.R., Falkingham P.L., Friedman M., Garwood R.J., Goswami A., Hutchinson J.R., Jeffery N.S.,
Johanson Z., Lebrun R., Martinez-Pérez C., Marugán-Lobón J., O’Higgins P.M., Metscher B., Orliac M.,
Rowe T.B., Rücklin M., Sánchez-Villagra M.R., Shubin N.H., Smith S.Y., Starck J.M., Stringer C.,
Summers A.P., Sutton M.D., Walsh S.A., Weisbecker V., Witmer L.M., Wroe S., Yin Z., Rayeld E.J. &
Donoghue P.C.J. 2017. Open data and digital morphology. Proceedings of the Royal Society B 284:
20170194. https://doi.org/10.1098/rspb.2017.0194
Davis G.R. & Elliott J.C. 1997. X-ray microtomography scanner using time-delay integration for
elimination of ring artefacts in the reconstructed image. Nuclear Instruments and Methods in Physics
Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 394 (1): 157–
162. https://doi.org/10.1016/S0168-9002(97)00566-4
Davit Y., Iltis G., Debenest G., Veran-Tissoires S., Wildenschild D., Gérino M. & Quintard M. 2011.
Imaging biolm in porous media using X-ray computed microtomography. Journal of Microscopy
242 (1): 15–25. https://doi.org/10.1111/j.1365-2818.2010.03432.x
de Carlo F., Gürsoy D., Ching D.J., Batenburg K.J., Ludwig W., Mancini L., Marone F., Mokso R.,
Pelt D.M., Sijbers J. & Rivers M. 2018. TomoBank: a tomographic data repository for computational x-ray
science. Measurement Science and Technology 29 (3): 034004. https://doi.org/10.1088/1361-6501/aa9c19
de Man B. 2001. Iterative Reconstruction for Reduction of Metal Artifacts in Computed Tomography.
PhD Thesis, KULeuven, Belgium.
Denison C., Carlson W.D. & Ketcham R.A. 1997. Three-dimensional quantitative textural analysis of
metamorphic rocks using high-resolution computed X-ray tomography: Part I. Methods and techniques.
Journal of Metamorphic Geology 15 (1): 29–44. https://doi.org/10.1111/j.1525-1314.1997.00006.x
European Journal of Taxonomy 522: 1–55 (2019)
38
DeVore M.L., Kenrick P., Pigg K.B. & Ketcham R.A. 2006. Utility of high resolution x-ray computed
tomography (HRXCT) for paleobotanical studies: an example using London Clay fruits and seeds.
American Journal of Botany 93 (12): 1848–1851. https://doi.org/10.3732/ajb.93.12.1848
Dhondt S., Vanhaeren H., Van Loo D., Cnudde V. & Inzé D. 2010. Plant structure visualization by high-
resolution X-ray computed tomography. Trends in Plant Science 15: 419–422.
https://doi.org/10.1016/j.tplants.2010.05.002
Dierick M., Cnudde V., Masschaele B., Vlassenbroeck J., Van Hoorebeke L. & Jacobs P. 2007.
Micro-CT of fossils preserved in amber. Nuclear Instruments and Methods in Physics Research
Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 580 (1): 641–643.
https://doi.org/10.1016/j.nima.2007.05.030
Dixon L.E., Bencivenga S. & Boden S.A. 2018. A new opening for wheat seed production. Journal of
Experimental Botany 69 (3): 341–343. https://doi.org/10.1093/jxb/erx430
Faraj K.A., Cuijpers V.M., Wismans R.G., Walboomers X.F., Jansen J.A., van Kuppevelt T.H. &
Daamen W.F. 2009. Micro-computed tomographical imaging of soft biological materials using contrast
techniques. Tissue Engineering Part C: Methods 15 (3): 493–499.
https://doi.org/10.1089/ten.tec.2008.0436
Faulwetter S., Vasileiadou A., Kouratoras M., Dailianis T. & Arvanitidis C. 2013a. Micro-computed
tomography: Introducing new dimensions to taxonomy. ZooKeys 263: 1–45.
https://doi.org/10.3897/zookeys.263.4261
Faulwetter S., Dailianis T., Vasileiadou A. & Arvanitidis C. 2013b. Contrast enhancing techniques for
the application of micro-CT in marine biodiversity studies. Microscopy and Analysis 27 (2): S4–S7.
Faulwetter S., Minadakis N., Keklikoglou K., Doerr M. & Arvanitidis C. 2015. First steps towards the
development of an integrated metadata management system for biodiversity-related micro-CT datasets.
Bruker microCT 2015 user meeting, Belgium: 149–159. Bruker, Bruges.
Fernández R., Kvist S., Lenihan J., Giribet G. & Ziegler A. 2014. Sine systemate chaos? A versatile tool
for earthworm taxonomy: non-destructive imaging of freshly xed and museum specimens using micro-
computed tomography. PloS ONE 9 (5): e96617. https://doi.org/10.1371/journal.pone.0096617
Franzen J.L., Gingerich P.D., Habersetzer J., Hurum J.H., von Koenigswald W. & Smith B.H. 2009.
Complete primate skeleton from the middle Eocene of Messel in Germany: morphology and paleobiology.
PLoS ONE 4 (5): e5723. https://doi.org/10.1371/journal.pone.0005723
Gamisch A., Staedler Y.M., Schönenberger J., Fischer G.A. & Comes H.P. 2013. Histological and micro-
CT evidence of stigmatic rostellum receptivity promoting auto-pollination in the Madagascan orchid
Bulbophyllum bicoloratum. PLoS ONE 8 (8): e72688. https://doi.org/10.1371/journal.pone.0072688
Garwood R.J., Sharma P.P., Dunlop J.A. & Giribet G. 2014. A Paleozoic stem group to mite harvestmen
revealed through integration of phylogenetics and development. Current Biology 24 (9): 1017–1023.
https://doi.org/10.1016/j.cub.2014.03.039
Gee C.T. 2013. Applying microCT and 3D visualization to Jurassic silicied conifer seed cones: a virtual
advantage over thin-sectioning. Applications in Plant Sciences 1 (11): 1300039.
https://doi.org/10.3732/apps.1300039
Gignac P.M. & Kley N.J. 2014. Iodine-enhanced micro-CT imaging: Methodological renements for
the study of the soft-tissue anatomy of post-embryonic vertebrates. Journal of Experimental Zoology
Part B: Molecular and Developmental Evolution 322 (3): 166–176. https://doi.org/10.1002/jez.b.22561
Gignac P.M., Kley N.J., Clarke J.A., Colbert M.W., Morhardt A.C., Cerio D., Cost I.N., Cox P.G.,
Daza J.D., Early C.M., Echolsm M.S., Henkelman R.M., Herdina A.N., Holliday C.M., Li Z.,
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
39
Mahlow K., Merchant S., Müller J., Orsbon C.P., Paluh D.J., Thies M.L., Tsai H.P. & Witmer L.M.
2016. Diffusible iodine-based contrast-enhanced computed tomography (diceCT): an emerging tool for
rapid, high-resolution, 3-D imaging of metazoan soft tissues. Journal of Anatomy 228 (6): 889–909.
https://doi.org/10.1111/joa.12449
Godfray H.C.J. 2007. Linnaeus in the information age. Nature 446: 259–260.
https://doi.org/10.1038/446259a
Golding R.E. & Jones A.S. 2007. Micro-CT as a novel technique for 3D reconstruction of molluscan
anatomy. Molluscan Research 27: 123–128.
Görög Á., Szinger B., Tóth E. & Viszkok J. 2012. Methodology of the micro-computer tomography on
Foraminifera. Palaeontologia Electronica 15 (1): 1–15.
Greco M.K., Welz P.M., Siegrist M., Ferguson S.J., Gallmann P., Roubik D.W. & Engel M.S. 2011.
Description of an ancient social bee trapped in amber using diagnostic radioentomology. Insectes
Sociaux 58 (4): 487–494. https://doi.org/10.1007/s00040-011-0168-8
Green O.R. 2001. A Manual of Practical Laboratory and Field Techniques in Palaeobiology. Springer
Science & Business Media, Berlin.
Grifn L.D., Elangovan P., Mundell A. & Hezel D.C. 2012. Improved segmentation of meteorite micro-
CT images using local histograms. Computers & Geosciences 39: 129-134.
https://doi.org/10.1016/j.cageo.2011.07.002
Hagedorn G., Mietchen D., Agosti D., Penev L., Berendsohn W. & Hobern D. 2011. Creative Commons
licenses and the non-commercial condition: Implications for the re-use of biodiversity information.
ZooKeys 150: 127–149. https://doi.org/10.3897/zookeys.150.2189
Hamba H., Nikaido T., Sadr A., Nakashima S. & Tagami J. 2012. Enamel lesion parameter correlations
between polychromatic micro-CT and TMR. Journal of Dental Research 91 (6): 586–591.
https://doi.org/10.1177/0022034512444127
Heřmanová Z., Dašková J., Ekrt B. & Kvaček J. 2017. Zlivifructus gen. nov., a new member of the
Normapolles complex. Review of Palaeobotany and Palynology 246: 177–184.
https://doi.org/10.1016/j.revpalbo.2017.06.012
Hezel D.C., Elangovan P., Viehmann S., Howard L., Abel R.L. & Armstrong R. 2013. Visualisation and
quantication of CV chondrite petrography using micro-tomography. Geochimica et Cosmochimica
Acta 116: 33–40. https://doi.org/10.1016/j.gca.2012.03.015
Hsieh J. 2009. Computed Tomography: Principles, Design, Artifacts, and Recent Advances. SPIE Press,
Bellingham, Washington. https://doi.org/10.1117/3.817303
Holliday C.M., Tsai H.P., Skiljan R.J., George I.D. & Pathan S. 2013. A 3D interactive model and atlas
of the jaw musculature of Alligator mississippiensis. PloS ONE 8 (6): e62806.
https://doi.org/10.1371/journal.pone.0062806
Jeppsson L., Anehus R. & Fredholm D. 1999. The optimal acetate buffered acetic acid technique for
extracting phosphatic fossils. Journal of Paleontology 73 (5): 964–972.
https://doi.org/10.1017/S0022336000040798
Johnson E., Dwarte D. & Jones A. 2011. Specimen preparation for X-ray micro-computed tomography
of plants. Microscopy and Microanalysis 17: 2010–2011. https://doi.org/10.1017/S1431927611002091
Kachelrieß M. 2008. Micro-CT. In: Semmler W. & Schwaiger M. (eds) Molecular Imaging I. Handbook
of Experimental Pharmacology 185/1: 23–52. Springer, Berlin / Heidelberg.
European Journal of Taxonomy 522: 1–55 (2019)
40
Kamenz C. & Weidemann G. 2009. Heavy metal—A contrasting substance for micro-tomographical
visualization of scorpion book lungs. Micron 40 (8): 911–917.
https://doi.org/10.1016/j.micron.2009.05.007
Kędzierski M., Uchman A., Sawlowicz Z. & Briguglio A. 2015. Fossilized bioelectric wire – the trace
fossil Trichichnus. Biogeosciences (Online) 12 (8): 2301–2309. https://doi.org/10.5194/bg-12-2301-2015
Kehlmaier C., Dierick M. & Skevington J.H. 2014. Micro-CT studies of amber inclusions reveal internal
genitalic features of big-headed ies, enabling a systematic placement of Metanephrocerus Aczél, 1948
(Insecta: Diptera: Pipunculidae). Arthopod Systematics & Phylogeny 72 (1): 23–36.
Keklikoglou K., Chatzinikolaou E., Faulwetter S., Grigoriou P., Papageorgiou N., Queirós A., Rühl S. &
Schmidbaur H. 2015. Revealing effects of ocean acidication on the calcied structures of marine
invertebrates through micro-computed tomography (micro-CT). Bruker microCT 2015 user meeting,
Belgium: 21–31. Bruker, Bruges.
Keklikoglou K., Faulwetter S., Chatzinikolaou E., Michalakis N., Filiopoulou I., Minadakis N.,
Panteri E., Perantinos G., Gougousis A. & Arvanitidis C. 2016. MicroCTvlab: A web based virtual gallery
of biological specimens using X-ray microtomography (microCT). Biodiversity Data Journal 4: e8740.
https://doi.org/10.3897/BDJ.4.e8740
Ketcham R.A. & Carlson W.D. 2001. Acquisition, optimization and interpretation of X-ray computed
tomographic imagery: applications to the geosciences. Computers & Geosciences 27 (4): 381–400.
https://doi.org/10.1016/S0098-3004(00)00116-3
Kim J.S., Min J., Recknagel A.K., Riccio M. & Butcher J.T. 2011. Quantitative three-dimensional
analysis of embryonic chick morphogenesis via microcomputed tomography. The Anatomical Record
294 (1): 1–10. https://doi.org/10.1002/ar.21276
Kim S.J., Xu W., Ahmad M.W., Baeck J.S., Chang Y., Bae J.E., Chae K.S., Jeong T.K., Park J.A. &
Lee G.H. 2015. Synthesis of nanoparticle CT contrast agents: in vitro and in vivo studies. Science and
Technology of Advanced Materials 16: 055003. https://doi.org/10.1088/1468-6996/16/5/055003
Knipfer T., Eustis A., Brodersen C., Walker A.M. & McElrone A.J. 2015. Grapevine species from varied
native habitats exhibit differences in embolism formation/repair associated with leaf gas exchange and
root pressure. Plant, Cell & Environment 38 (8): 1503–1513. https://doi.org/10.1111/pce.12497
Kniss J., Kindlmann G. & Hansen C. 2002. Multidimensional transfer functions for interactive
volume rendering. IEEE Transactions on Visualization and Computer Graphics 8 (3): 270–285.
https://doi.org/10.1109/TVCG.2002.1021579
Koenigswald W.V., Habersetzer J. & Gingerich P.D. 2011. Morphology and evolution of the distal
phalanges in primates. In: Lehmann T. & Schaal S.F.K (eds) The World at the Time of Messel: Puzzles
in Palaeobiology, Palaeoenvironment, and the History of Early Primates: 91–94. 22nd International
Senckenberg Conference, Frankfurt am Main.
Konietzko-Meier D. & Schmitt A. 2013. A histological study of a femur of Plagiosuchus, a Middle
Triassic temnospondyl amphibian from southern Germany, using thin sections and micro-CT scanning.
Netherlands Journal of Geosciences 92 (2–3): 97–108. https://doi.org/10.1017/S0016774600000020
Korte N. & Porembski S. 2011. Anatomical analysis of turgescent and semi-dry resurrection plants:
the effect of sample preparation on the sample, resolution, and image quality of X-ray micro-computed
tomography (μCT). Microscopy Research and Technique 74 (4): 364–369.
https://doi.org/10.1002/jemt.20917
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
41
Krings M., Müller H., Heneka M.J. & Rödder D. 2017. Modern morphological methods for tadpole
studies. A comparison of micro-CT, and clearing and staining protocols modied for frog larvae.
Biotechnic & Histochemistry 92 (8): 595–605. https://doi.org/10.1080/10520295.2017.1369162
Kuebler K.E., McSween H.Y., Carlson W.D. & Hirsch D. 1999. Sizes and masses of chondrules and
metal–troilite grains in ordinary chondrites: possible implications for nebular sorting. Icarus 141 (1):
96–106. https://doi.org/10.1006/icar.1999.6161
Kuva J., Sammaljärvi J., Parkkonen J., Siitari-Kauppi M., Lehtonen M., Turpeinen T., Timonen J. &
Voutilainen M. 2018. Imaging connected porosity of crystalline rock by contrast agent-aided X-ray
microtomography and scanning electron microscopy. Journal of Microscopy 270 (1): 98–109.
https://doi.org/10.1111/jmi.12661
Kvaček J. & Heřmanová Z. 2017. Fossil angiosperm fruit Allericarpus parvivalvis (Ericales) from
the Coniacian of the Bohemian Cretaceous Basin. Journal of the National Museum (Prague), Natural
History Series 186: 117–124.
Kvaček J., Doyle J.A., Endress P.K., Daviero-Gomez V., Gomez B. & Tekleva M. 2016.
Pseudoasterophyllites cretaceus from the Cenomanian (Cretaceous) of the Czech Republic – a possible
link between Chloranthaceae and Ceratophyllum. Taxon 65 (6): 1345–1373.
https://doi.org/10.12705/656.8
Kvaček J., Barrón E., Heřmanová Z., Mendes M.M., Karch J., Žemlička J. & Dudák J. 2018a.
Araucarian conifer from late Albian amber of northern Spain. Papers in Palaeontology 4 (3): 1–14.
https://doi.org/10.1002/spp2.1223
Kvaček J., Heřmanová Z., Bruthansová J.M.M., Karch J., Žemlička J. & Dudák J. 2018b. Stutzeliastrobus
bohemicus - basal Curessaceae conifer from the Cenomanian of the Bohemian Cretaceous Basin, Central
Europe. Fossil Imprint 74: 179–188.
Leroux O., Leroux F., Bellefroid E., Claeys M., Couvreur M., Borgonie G., Van Hoorebeke L.,
Masschaele B. & Viane R. 2009. A new preparation method to study fresh plant structures with X-ray
computed tomography. Journal of Microscopy 233 (1): 1–4.
https://doi.org/10.1111/j.1365-2818.2008.03088.x
Li X., Pan Z., Upadhyaya S.K., Atungulu G.G. & Delwiche M. 2011. Three-dimensional geometric
modeling of processing tomatoes. Transactions of the American Society of Agricultural and Biological
Engineers 54 (6): 2287–2296. https://doi.org/10.13031/2013.40642
Li Z., Clarke J.A., Ketcham R.A., Colbert M.W. & Yan F. 2015. An investigation of the efcacy and
mechanism of contrast-enhanced X-ray computed tomography utilizing iodine for large specimens
through experimental and simulation approaches. BMC Physiology 15: 5.
https://doi.org/10.1186/s12899-015-0019-3
Liu Y., Scholtz G. & Hou X. 2015. When a 520 million-year-old Chengjiang fossil meets a modern
micro-CT–a case study. Scientic Reports 5: 12802. https://doi.org/10.1038/srep12802
Lorensen W.E. & Cline H.E. 1987. Marching cubes: a high resolution 3D surface construction algorithm.
ACM Siggraph Computer Graphics 21 (4): 163–169.
Mancini L., Kourousias G., Billè F., De Carlo F. & Fidler A. 2018. About a method for compressing
x-ray computed microtomography data. Measurement Science and Technology 29 (4): 044002.
https://doi.org/10.1088/1361-6501/aaa0fa
McElrone A.J., Choat B., Parkinson D.Y., MacDowell A.A. & Brodersen C.R. 2013. Using high
resolution computed tomography to visualize the three dimensional structure and function of plant
vasculature. Journal of Visualized Experiments 74: e50162. https://doi.org/10.3791/50162
European Journal of Taxonomy 522: 1–55 (2019)
42
Meganck J.A., Kozloff K.M., Thornton M.M., Broski S.M. & Goldstein S.A. 2009. Beam hardening
artifacts in micro-computed tomography scanning can be reduced by X-ray beam ltration
and the resulting images can be used to accurately measure BMD. Bone 45 (6): 1104–1116.
https://doi.org/10.1016/j.bone.2009.07.078
Metscher B.D. 2009a. MicroCT for comparative morphology: simple staining methods allow
high-contrast 3D imaging of diverse non-mineralized animal tissues. BMC Physiology 9: 11.
https://doi.org/10.1186/1472-6793-9-11
Metscher B.D. 2009b. MicroCT for developmental biology: A versatile tool for high-contrast 3D imaging
at histological resolutions. Developmental Dynamics 238 (3): 632–640.
https://doi.org/10.1002/dvdy.21857
Metscher B.D. 2011. X-ray microtomographic imaging of intact vertebrate embryos. Cold Spring
Harbor Protocols 12: 1462–1471. https://doi.org/10.1101/pdb.prot067033
Meyer M., Elliott D., Wood A.D., Polys N.F., Colbert M., Maisano J.A., Vickers-Rich P., Hall M.,
Hoffman K.H., Schneider G. & Xiao S. 2014. Three-dimensional microCT analysis of the Ediacara fossil
Pteridinium simplex sheds new light on its ecology and phylogenetic afnity. Precambrian Research
249: 79–87. https://doi.org/10.1016/j.precamres.2014.04.013
Miller J.S. & Venable D.L. 2003. Floral morphometrics and the evolution of sexual dimorphism in
Lycium (Solanaceae). Evolution 57 (1): 74–86. https://doi.org/10.1111/j.0014-3820.2003.tb00217.x
Mizutani R. & Suzuki Y. 2012. X-ray microtomography in biology. Micron 43 (2–3): 104–115.
https://doi.org/10.1016/j.micron.2011.10.002
Mizutani R., Takeuchi A., Hara T., Uesugi K. & Suzuki Y. 2007. Computed tomography imaging
of the neuronal structure of Drosophila brain. Journal of Synchrotron Radiation 14 (3): 282–287.
https://doi.org/10.1107/S0909049507009004
Mizutani R., Takeuchi A., Akamatsu G., Uesugi K. & Suzuki Y. 2008a. Element-specic microtomo-
graphic imaging of Drosophila brain stained with high-Z probes. Journal of Synchrotron Radiation 15
(4): 374–377. https://doi.org/10.1107/S0909049508003725
Mizutani R., Takeuchi A., Uesugi K., Takekoshi S., Osamura R.Y. & Suzuki Y. 2008b. X-ray micro-
tomographic imaging of three-dimensional structure of soft tissues. Tissue Engineering Part C: Methods
14 (4): 359–363. https://doi.org/10.1089/ten.tec.2008.0274
Mizutani R., Takeuchi A., Uesugi K., Takekoshi S., Osamura R.Y. & Suzuki Y. 2009. Three-dimensional
microstructural analysis of human brain tissue by using synchrotron radiation microtomographs. In:
Westland T.B. & Calton R.N. (eds) Handbook on White Matter: 247–277. Nova Science Publishers,
New York.
Mooney S.J., Pridmore T.P., Helliwell J. & Bennett M.J. 2012. Developing X-ray computed
tomography to non-invasively image 3-D root systems architecture in soil. Plant and Soil 352: 1–22.
https://doi.org/10.1007/s11104-011-1039-9
Needham A.W., Abel R.L., Tomkinson T. & Grady M.M. 2013. Martian subsurface uid pathways
and 3D mineralogy of the Nakhla meteorite.Geochimica et Cosmochimica Acta 116: 96-110.
https://doi.org/10.1016/j.gca.2012.07.004
Pajor R., Fleming A., Osborne C.P., Rolfe S.A., Sturrock C.J. & Mooney S.J. 2013. Seeing space:
visualization and quantication of plant leaf structure using X-ray micro-computed tomography: View
Point. Journal of Experimental Botany 64 (2): 385–390. https://doi.org/10.1093/jxb/ers392
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
43
Paquit V.C., Gleason S.S. & Kalluri U.C. 2011. Monitoring plant growth using high
resolution micro-CT images. Proceedings of SPIE-IS&T Electronic Imaging 7877: 78770W.
https://doi.org/10.1117/12.876719
Paredes U.M., Prys-Jones R., Adams M., Groombridge J., Kundu S., Agapow P.M. & Abel R.L. 2012.
Micro-CT X-rays do not fragment DNA in preserved bird skins. Journal of Zoological Systematics and
Evolutionary Research 50 (3): 247–250. https://doi.org/10.1111/j.1439-0469.2012.00657.x
Paterson G.L., Sykes D., Faulwetter S., Merk R., Ahmed F., Hawkins L.E., Dinley J., Bal A.D. &
Arvanitidis C. 2014. The pros and cons of using micro-computed tomography in gross and micro-
anatomical assessments of polychaetous annelids. Memoirs of Museum Victoria 71: 237–46.
Pauwels E., Van Loo D., Cornillie P., Brabant L. & Van Hoorebeke L. 2013. An exploratory study of
contrast agents for soft tissue visualization by means of high resolution X-ray computed tomography
imaging. Journal of Microscopy 250 (1): 21–31. https://doi.org/10.1111/jmi.12013
Penney D., Dierick M., Cnudde V., Masschaele B., Vlassenbroeck J., Van Hoorebeke L. & Jacobs P.
2007. First fossil Micropholcommatidae (Araneae), imaged in Eocene Paris amber using X-ray computed
tomography. Zootaxa 1623: 47–53. https://doi.org/10.5281/zenodo.179249
Plougonven E., Srichamnong W., Srzednicki G. & Léonard A. 2012. Characterisation of macadamia nuts
using X-ray microtomography. SkyScan MicroCT User Meeting 2012, Belgium: 1–5. Skyscan, Brussels.
Presnell J.K. & Schreibman M.P. 1997. Humason’s Animal Tissue Techniques. Fifth edition. Johns
Hopkins University Press, Baltimore / London.
Rahman I.A. & Smith S.Y. 2014. Virtual paleontology: computer-aided analysis of fossil form and
function. Journal of Paleontology 88 (4): 633–635. https://doi.org/10.1666/13-001I
Rivers M.L., Sutton S.R. & Eng P.J. 1999. Geoscience applications of X-ray computed
microtomography. In: Developments in X-Ray Tomography II (Vol. 3772): 78–86. SPIE’s International
Symposium on Optical Science, Engineering, and Instrumentation, Denver, CO, United States.
https://doi.org/10.1117/12.363741
Roche R.C., Abel R.A., Johnson K.G. & Perry C.T. 2010. Quantication of porosity in Acropora pulchra
(Brook 1891) using X-ray micro-computed tomography techniques. Journal of Experimental Marine
Biology and Ecology 396 (1): 1–9. https://doi.org/10.1016/j.jembe.2010.10.006
Rossi M., Casali F., Romani D., Bondioli L., Macchiarelli R. & Rook L. 2004. MicroCT scan in
paleobiology: application to the study of dental tissues. Nuclear Instruments and Methods in Physics
Research Section B: Beam Interactions with Materials and Atoms 213: 747–750.
https://doi.org/10.1016/S0168-583X(03)01697-5
Ruthensteiner B., Baeumler N. & Barnes D.G. 2010. Interactive 3D volume rendering in biomedical
publications. Micron 41 (7): 886-e1. https://doi.org/10.1016/j.micron.2010.03.010
Salmon P.L., Liu X. & Sasov A. 2009. A post-scan method for correcting artefacts of slow geometry
changes during micro-tomographic scans. Journal of X-Ray Science and Technology 17 (2): 161–174.
Sansone S., Rocca-Serra P., Field D., Maguire E., Taylor C., Hofmann O., Fang H., Neumann S., Tong
W., Amaral-Zettler L., Begley K., Booth T., Bougueleret L., Burns G., Chapman B., Clark T., Coleman
L., Copeland J., Das S., de Daruvar A., de Matos P., Dix I., Edmunds S., Evelo C.T., Forster M.J.,
Gaudet P., Gilbert J., Goble C., Grifn J.L., Jacob D., Kleinjans J., Harland L., Haug K., Hermjakob
H., Sui S.J.H., Laederach A., Liang S., Marshall S., McGrath A., Merrill E., Reilly D., Roux M., Shamu
C.E., Shang C.A., Steinbeck C., Trefethen A., Williams-Jones B., Wolstencroft K., Xenarios I. & Hide
W. 2012. Toward interoperable bioscience data. Nature Genetics 44 (2): 121–126.
https://doi.org/10.1038/ng.1054
European Journal of Taxonomy 522: 1–55 (2019)
44
Schambach S.J., Bag S., Schilling L., Groden C. & Brockmann M.A. 2010. Application of micro-CT in
small animal imaging. Methods 50 (1): 2–13. https://doi.org/10.1016/j.ymeth.2009.08.007
Schmidbaur H., Keklikoglou K., Metscher B.D. & Faulwetter S. 2015. Exploring methods to remove
iodine and phosphotungstic acid stains from zoological specimens. Micro-CT User Meeting, Belgium:
116–123. Bruker, Bruges.
Simons F.J., Verhelst F. & Swennen R. 1997. Quantitative characterization of coal by means of microfocal
X-ray computed microtomography (CMT) and color image analysis (CIA). International Journal of
Coal Geology 34 (1): 69–88. https://doi.org/10.1016/S0166-5162(97)00011-6
Staedler Y.M., Masson D. & Schönenberger J. 2013. Plant tissues in 3D via X-ray tomography: simple
contrasting methods allow high resolution imaging. PLoS ONE 8 (9): e75295.
https://doi.org/10.1371/journal.pone.0075295
Staedler Y.M., Kreisberger T., Manafzadeh S., Chartier M., Handschuh S., Pamperl S., Sontag S.,
Paun O. & Schönenberger J. 2018. Novel computed tomography-based tools reliably quantify plant
reproductive investment. Journal of Experimental Botany 69 (3): 525–535.
https://doi.org/10.1093/jxb/erx405
Stanley S.M. 1992. Exploring Earth and Life Through Time. W H Freeman & Co, New York.
Stoev P., Komerički A., Akkari N., Liu S., Zhou X., Weigand A.M., Hostens J., Porco D.& Penev L. 2013.
Transcriptomic, DNA barcoding, and micro-CT imaging data from an advanced taxonomic description
of a novel centipede species (Eupolybothrus cavernicolus Komerički & Stoev, sp. n.). GigaScience.
https://doi.org/10.5524/100063
Stuppy W.H., Maisano J.A., Colbert M.W., Rudall P.J. & Rowe T.B. 2003. Three-dimensional analysis
of plant structure using high-resolution X-ray computed tomography. Trends in Plant Science 8 (1): 2–6.
https://doi.org/10.1016/S1360-1385(02)00004-3
Sutton M.D. 2008. Tomographic techniques for the study of exceptionally preserved fossils. Proceedings
of the Royal Society of London B 275 (1643): 1587–1593. https://doi.org/10.1098/rspb.2008.0263
Sutton M.D., Briggs D.E., Siveter D.J. & Siveter D.J. 2001. Methodologies for the visualization and
reconstruction of three-dimensional fossils from the Silurian Herefordshire Lagerstätte. Palaeontologia
Electronica 4 (1): 1–17.
Sutton M., Rahman I. & Garwood R. 2014. Techniques for Virtual Palaeontology. John Wiley & Sons,
New York.
Vågberg W., Larsson J.C. & Hertz H.M. 2017. Removal of ring artifacts in microtomography by
characterization of scintillator variations. Optics Express 25 (19): 23191–23198.
https://doi.org/10.1364/OE.25.023191
van der Niet T., Zollikofer C.P.E., de León M.S.P., Johnson S.D. & Linder H.P. 2010. Three-dimensional
geometric morphometrics for studying oral shape variation. Trends in Plant Science 15 (8): 423–426.
https://doi.org/10.1016/j.tplants.2010.05.005
van Soldt B.J., Metscher B.D., Poelmann R.E., Vervust B., Vonk F.J., Muller G.B. & Richardson M.K.
2015. Heterochrony and early left-right asymmetry in the development of the cardiorespiratory system
of snakes. PLoS ONE 10 (1): e116416. https://doi.org/10.1371/journal.pone.0116416
Vodrážka R. 2009. A new method for the extraction of macrofossils from calcareous rocks using sulphuric
acid. Palaeontology 52 (1): 187–192. https://doi.org/10.1111/j.1475-4983.2008.00829.x
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
45
Walter T., Shattuck D.W., Baldock R., Bastin M.E., Carpenter A.E., Duce S., Ellenberg J., Fraser A.,
Hamilton N., Pieper S., Ragan M.A., Schneider J.E., Tomancak P. & Hériché J.-K. 2010. Visualization
of image data from cells to organisms. Nature Methods 7: S26–S41.
https://doi.org/10.1038/nmeth.1431
Wan B., Xiao S., Yuan X., Chen Z., Pang K., Tang Q., Guan C. & Maisano J.A. 2014. Orbisiana
linearis from the early Ediacaran Lantian Formation of South China and its taphonomic and ecological
implications. Precambrian Research 255: 266–275. https://doi.org/10.1016/j.precamres.2014.09.028
Wang C.N., Hsu H.C., Wang C.C., Lee T.K. & Kuo Y.F. 2015. Quantifying oral shape variation in 3D
using microcomputed tomography: a case study of a hybrid line between actinomorphic and zygomorphic
owers. Frontiers in Plant Science 6: 724. https://doi.org/10.3389/fpls.2015.00724
Wellington S.L. & Vinegar H.J. 1987. X-ray computerized tomography. Journal of Petroleum Technology
39: 885–898.
Wellman C. & Axe L. 1999. Extracting plant mesofossils and megafossils by bulk acid maceration. In:
Jones T.P. & Rowe N.P. (eds) Fossil Plants and Spores: Modern Techniques: 11–14. Geological Society,
London.
Yakami M., Ishizu K., Kubo T., Okada T. & Togashi K. 2011. Development and evaluation of a low-cost
and high-capacity DICOM image data storage system for research. Journal of Digital Imaging 24 (2):
190–195. https://doi.org/10.1007/s10278-009-9267-8
Yang H., Chen Z.Q., Wang Y., Ou W., Liao W. & Mei X. 2015. Palaeoecology of microconchids from
microbialites near the Permian–Triassic boundary in South China. Lethaia 48 (4): 497–508. https://doi.
org/10.1111/let.12122
Zinsmeister W.J. & De Nooyer C. 1996. Computed tomography; non-destructive techniques for
visualizing internal morphology of invertebrates. Geological Society of America, Abstracts with
Programs 28 (7): 294.
Manuscript received: 12 September 2018
Manuscript accepted: 21 January 2019
Published on: 30 April 2019
Topic editor: Koen Martens
Desk editor: Kristiaan Hoedemakers
Printed versions of all papers are also deposited in the libraries of the institutes that are members of
the EJT consortium: Muséum national d’Histoire naturelle, Paris, France; Meise Botanic Garden,
Belgium; Royal Museum for Central Africa, Tervuren, Belgium; Royal Belgian Institute of Natural
Sciences, Brussels, Belgium; Natural History Museum of Denmark, Copenhagen, Denmark; Naturalis
Biodiversity Center, Leiden, the Netherlands; Museo Nacional de Ciencias Naturales-CSIC, Madrid,
Spain; Real Jardín Botánico de Madrid CSIC, Spain; Zoological Research Museum Alexander Koenig,
Bonn, Germany.
European Journal of Taxonomy 522: 1–55 (2019)
46
Appendix
1. Glossary of terms
A glossary of terms related with micro-CT technology and imaging has been compiled and presented in
Table 11.
2. List of institutions hosting a micro-CT
A comprehensive list of academic and research institutions operating micro-CT systems is presented in
Table 12 for SYNTHESYS3 partners (EU FP7 programme) and in Table 13 for collaborators all over
the world.
3. List of micro-CT manufacturers
A list of micro-CT manufacturers and their company links is presented in Table 14.
4. List of software
An up-to-date list of software currently used for post-processing procedures is presented in Table 7 for
3D volume rendering, in Table 8 for segmentation, in Table 9 for 2D/3D analysis and in Table 15 for 2D
visualisation.
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
47
3D Three-dimensional
CCD camera Charge-coupled device camera: a technology used in digital photography to
convert a light signal into a digital value.
Creative Commons licences A set of copyright-licences that describe which rights of the licenced material
are reserved and to what extent and under which condition the material can be
modied and shared. The licences are standardised, simple to understand and
computer-readable (http://www.creativecommons.org).
CT Computer tomography, a non-destructive imaging technique that converts
projection views into orthogonal slices to calculate three-dimensional
representations of the imaged object.
CTVox A commercial volume rendering software developed by SkyScan / Bruker.
DICOM Digital Imaging and Communications in Medicine: a standard providing
format denitions and a communication protocol for the description and
exchange of medical image data.
DOI Digital Object Identier: a digital identier used to uniquely identify an object
and retrieve metadata about it. Mostly applied to documents such as scientic
papers, books, reports, etc. but not restricted to these – any digital object can
be assigned a DOI.
Drishti Free volume rendering software (http://sf.anu.edu.au/Vizlab/drishti).
Dryad / Datadryad An online data repository to store data underlying scientic publications.
Assigns DOIs to the submitted data so that they can be discovered and cited
(http://datadryad.org).
gshare An online data repository to store data underlying scientic publications.
Assigns DOIs to the submitted data so that they can be discovered and cited
(http://www.gshare.com).
Flat-eld correction A calibration procedure to improve the images created by the CCD camera.
During the at-eld correction, defect pixels and camera artefacts are
accounted for, and thus they can be removed in the nal image.
GB Gigabyte: 109 bytes. A byte is a unit of digital information, its multiples
(kilo-, mega-, gigabyte) are often used to describe the size of a digital storage
medium, le or dataset.
GigaDB An online data repository to store data underlying scientic publications,
specializing on very large datasets. Assigns DOIs to the submitted data so that
they can be discovered and cited (http://www.gigadb.org).
HMDS Hexamethyldisilazane: a chemical used for drying specimens without affecting
their morphology.
ISA-TAB A newly proposed data format intended to facilitate integration of diverse
datasets across various life-science disciplines (https://isa-tools.org/).
isosurface model Geometrical, three-dimensional model consisting of surfaces of equal densities
(often triangles).
keV Kilo electron volt
Table 11 (continued on next page). Glossary of terms.
European Journal of Taxonomy 522: 1–55 (2019)
48
kV Kilovolt
kVp Peak kilovoltage
metadata Data about data. Ambiguous term, since data can be both “normal data” as
well as “metadata”, depending on the context. Metadata help to understand the
context of the data as well as provide a means to integrate it with other data.
µA Micro-ampere
morphometric analysis Statistical analysis based on the shape and size of morphological characters
of organisms to detect similarities or signicant differences between samples,
populations or species.
OBJ Wavefront Object: a data format used to represent three-dimensional geometric
objects with the information of colour.
pixel The smallest unit of information in a digital image.
projection image Also called shadow image or radiograph. Image that results from CT
imaging provides a grayscale side view of the specimen, with dense parts in
darker shades and less dense parts in lighter shades of grey. From a series of
projection images the cross-sections are calculated.
PTA Phosphotungstic acid: an electron-dense stain used as a contrast agent in
histology and CT imaging.
RAM Random-access memory: a form of computer data storage for fast access.
reconstruction Algorithmic process of transforming projection images into cross sections.
ROI Region of Interest: User-specied region of an image.
SEM Scanning Electron Microscopy: a high-resolution imaging technique.
surface rendering, surface
model
See isosurface model.
transfer values, transfer
function
In volume rendering, a denition of the colour and opacity values with which
each voxel should be displayed.
type material, type specimen A biological reference specimen, usually kept in museums or biological
collections, to which the scientic name of that taxon is formally attached.
URI Uniform Resource Identier: a sequence of characters used to identify a web
resource.
URL Uniform Resource Locator: often also synonymously used with ‘web address’,
a URL is a URI which also provides a means to access the resource.
volume rendering A way of displaying three-dimensional data acquired through 3D-imaging
techniques such as CT or MRI. Depending on the transfer functions applied,
different densities of the sample can be visualised and optionally colour-coded.
voxel Volume pixel: the smallest unit of information in a three-dimensional digital
space.
Table 11 (continued).
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
49
Table 12. Overview of the micro-CT labs hosted by the SYNTHESYS3 partners.
Micro-CT lab Country Micro-CT
system
Further information
Natural History
Museum in London
United
Kingdom
Nikon
Metrology
HMX ST
225
http://www.nhm.ac.uk/our-science/departments-and-staff/core-research-labs/imaging-and-analysis/computed-tomography.html
Institute of Marine
Biology and
Genetics (HCMR)
Greece Skyscan
1172
https://microct.portal.lifewatchgreece.eu/
Muséum national
d’Histoire naturelle
(MNHN)
France GE
Phoenix
VTomeX
L240
http://ums2700.mnhn.fr/ast-rx/acces#tarication
Museo nacional de
Ciencias naturales
(MNCN)
Spain Nikon XT-
H-160
http://www.mncn.csic.es/index.jsp?seccion=1325&id=2010120902260001&activo=12
Wageningen
University and
Research Centre
Netherlands GE
Phoenix
v[tome] x
m
http://www.wageningenur.nl/en/Expertise-Services/Facilities/CATAgro
Food-3/CATAgroFood-3/Our-facilities/Show/Xray-computer-Tomography-
XRT-CT-Noninvasive-3D-imaging-of-internal-structures.htm
Naturalis
Biodiversity Center
Netherlands Skyscan
1172
https://science.naturalis.nl/en/labs-services/laboratories/geolab/
Národní Museum Czech
Republic
Skyscan
1172
http://www.nm.cz/
Museum für
Naturkunde
Germany Phoenix
Nanotome|s
https://muellerlaboratory.wordpress.com/
https://www.kristin-mahlow.de/mico-ct-labor/
Royal Belgian
Institute of Natural
Sciences
Belgium RX
EasyTom
XRE
UniTom
http://collections.naturalsciences.be
http://virtualcollections.naturalsciences.be/
European Journal of Taxonomy 522: 1–55 (2019)
50
Micro-CT lab Country Micro-CT
system
Further information
Natural History
Museum in London
United
Kingdom
Nikon
Metrology
HMX ST
225
http://www.nhm.ac.uk/our-science/departments-and-staff/core-research-labs/imaging-and-analysis/computed-tomography.html
Institute of Marine
Biology and
Genetics (HCMR)
Greece Skyscan
1172
https://microct.portal.lifewatchgreece.eu/
Muséum national
d’Histoire naturelle
(MNHN)
France GE
Phoenix
VTomeX
L240
http://ums2700.mnhn.fr/ast-rx/acces#tarication
Museo nacional de
Ciencias naturales
(MNCN)
Spain Nikon XT-
H-160
http://www.mncn.csic.es/index.jsp?seccion=1325&id=2010120902260001&activo=12
Wageningen
University and
Research Centre
Netherlands GE
Phoenix
v[tome] x
m
http://www.wageningenur.nl/en/Expertise-Services/Facilities/CATAgro
Food-3/CATAgroFood-3/Our-facilities/Show/Xray-computer-Tomography-
XRT-CT-Noninvasive-3D-imaging-of-internal-structures.htm
Naturalis
Biodiversity Center
Netherlands Skyscan
1172
https://science.naturalis.nl/en/labs-services/laboratories/geolab/
Národní Museum Czech
Republic
Skyscan
1172
http://www.nm.cz/
Museum für
Naturkunde
Germany Phoenix
Nanotome|s
https://muellerlaboratory.wordpress.com/
https://www.kristin-mahlow.de/mico-ct-labor/
Royal Belgian
Institute of Natural
Sciences
Belgium RX
EasyTom
XRE
UniTom
http://collections.naturalsciences.be
http://virtualcollections.naturalsciences.be/
Table 13 (continued on next two pages). Indicative list of micro-CT labs all over the world.
Micro-CT labs Country Micro-CT system Further information
Henry Moseley X-ray
imaging facility, University of
Manchester
United
Kingdom
E.g., Nikon Metrology
225/320 kV
Nikon XTH 225
http://www.dalton.manchester.ac.uk/research/facilities/manchesterx-rayimagingfacility/
The Musculoskeletal Imaging
Core at UMASS Medical
School
USA Scanco MicroCT40 https://www.umassmed.edu/radiology/clinical-divisions/musculoskeletal-imaging/welcome/
CNS - Center for Nanoscale
Systems
USA Nikon Metrology HMX
ST 225
https://cns1.rc.fas.harvard.edu/
Rush University Medical
Center - microCT Core Lab
USA Scanco MicroCT40
Scanco MicroCT50
https://www.rushu.rush.edu/research/rush-core-laboratories/rush-microct-and-histology-core
University of Minnesota XRCT
lab
USA X-View CT X5000 http://xraylab.esci.umn.edu/
High-Resolution X-ray
Computed Tomography Facility
at the University of Texas at
Austin (UTCT)
USA Xradia microXCT 400
NSI scanner
http://www.ctlab.geo.utexas.edu/
University of Mississippi
Medical Center Department of
Biomedical Materials Science
USA Skyscan 1172 https://www.umc.edu/sod/
Centre for X-ray Tomography
of Ghent University (UGCT)
Belgium NanoWood
Environmental Micro-CT
Medusa
http://www.ugct.ugent.be/
Institute for Biomechanics -
ETH Zürich
Switzerland Scanco µCT 41, 50 and
80
http://www.biomech.ethz.ch/
Mario Negri Institute of Milan Italy GE Explore Locus http://www.marionegri.it/en_US/home
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
51
Micro-CT labs Country Micro-CT system Further information
Weizmann Institute of Science Israel Xradia microXCT 400 http://www.weizmann.ac.il
Stellenbosch University CT
Scanner Unit
South Africa GE Phoenix VTomeX
L240
https://www.sun.ac.za/english/faculty/science/CAF/units/ct-scanner
Aix-Marseille Université
(AMU), Laboratoire Cerege
France MicroXCT-400 Zeiss-
Xradia
https://www.nano-id.fr/tomographie-x3d/
De la Prehistoire à l’ Actuel:
Culture, Environnement,
Anthropologie (PACEA)
France GE Phoenix VTomeX http://www.pacea.u-bordeaux1.fr/
Laboratoire Biogéosciences,
University of Bourgogne
France SkyScan 1174 http://biogeosciences.u-bourgogne.fr/en/
Institut de Paléoprimatologie,
Paléontologie
Humaine: Evolution et
Paléoenvironnements (IPHEP),
Université de Poitiers
France RX solutions EasyTom
XL Duo
Viscom X8050
http://sfa.univ-poitiers.fr/imageup/equipements/materiel-dans-le-laboratoire-ic2mp/?post_type=social
Structure Fédérative de
Recherche de Lyon Gerland
(SFR), BioSciences
France GE Phoenix Nanotom S http://www.sfr-biosciences.fr/
http://www.sfr-biosciences.fr/plateformes/Phenotypage-cellulaire/AniRA-ImmOs/AniRA-immos
Laboratoire IPANEMA France Psiché
Metrologie
Anatomix
http://www.synchrotron-soleil.fr/Recherche/LignesLumiere/PSICHE
http://www.synchrotron-soleil.fr/Recherche/LignesLumiere/METROLOGIE
http://ipanema.cnrs.fr/spip/
MRI-ISEM, Evolutionary
Institute, University of
Montpellier
France EasyTom 150 http://www.mri.cnrs.fr/en/x-ray-tomography/mri-isem.html
Table 13 (continued).
European Journal of Taxonomy 522: 1–55 (2019)
52
Micro-CT labs Country Micro-CT system Further information
Theoretical Biology MicroCT
Imaging Lab, University of
Vienna
Austria MicroXCT-200
Zeiss-Xradia
Skyscan 1174
https://theoretical.univie.ac.at/microct-lab/
Vienna Micro-CT Lab,
Department of Anthropology
Austria Viscom X8060 http://www.micro-ct.at/en/
Department of Palaeontology,
University of Vienna
Austria Skyscan 1173 http://www.univie.ac.at/Palaeontologie/FACILITIES2_DE.html
Institute of Biotechnology,
Cornell University
USA Xradia Zeiss
VersaXRM-520
GE eXplore CT-120
http://www.biotech.cornell.edu/node/573
Table 13 (continued).
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
53
Table 14. MicroCT manufacturers and their company links (modied table from Schambach et al. 2010
and http://microctworld.net/).
Company Website
Bruker (acquisition of Skyscan and Carestream) http://bruker-microct.com/home.htm
http://www.carestream.com/default.aspx?LangType=1033
CT imaging http://www.ct-imaging.de/en/
Echo Medical Systems http://www.echomri.com/
FEI (ThermoFisher) http://www.fei.com/
GE Medical Systems http://www.gehealthcare.com
GE Measurement and Control Solutions https://www.industrial.ai/
Hitachi Aloka Medical http://www.hitachi-aloka.co.jp/english/
Mediso http://www.mediso.hu/
Milabs http://www.milabs.com/
NanoFocus Ray http://www.nfr.kr/e_index.php
Nikon Metrology http://www.nikonmetrology.com/en_EU/
North Star Imaging (XviewCT) http://www.xviewct.com/
Perkin Elmer http://www.perkinelmer.com/
Prexion http://www.prexion.com/
RX Solutions http://www.rxsolutions.fr/
Scanco Medical http://www.scanco.ch/
Sedecal https://www.sedecal.com/en/
Siemens http://www.siemens.com/entry/cc/en/
Stratec Medizintechnik GmbH http://www.stratec-med.com
Toshiba IT & Control Systems Corporation https://www.toshiba-itc.com/en/
TriFoil imaging (acquisition of Bioscan Inc.) http://www.trifoilimaging.com/
Varian Medical Systems https://www.varian.com
Werth Messtechnik http://www.werth.de/de/start/home.html
Xstrahl http://www.xstrahl.com/
XRE https://xre.be/
YXLON International GmbH http://www.yxlon.com/
Zeiss http://www.xradia.com/
European Journal of Taxonomy 522: 1–55 (2019)
54
Table 15. Software for 2D visualisation (modied table from Walter et al. 2010 and Abel et al. 2012).
Software Licence Type URL
Amira Commercial www.amira.com
Arivis (web-based software) Commercial http://vision.arivis.com/
BioImageXD Free http://www.bioimagexd.net
Brain Maps (web-based software) Free http://brainmaps.org
Dragony
Free licences available
for researchers with non-
commercial activities/
Commercial
http://www.theobjects.com/dragony/
Fiji (Is Just ImageJ) Free http://ji.sc/
Huygens Commercial http://www.svi.nl
ImageJ Free https://imagej.nih.gov/ij/
Image-Pro Commercial http://www.mediacy.com
Imaris Commercial http://www.bitplane.com/
Mimics Commercial www.materialise.com/mimics
Octopus Commercial https://octopusimaging.eu/
Simpleware Commercial www.simpleware.com
Slice:Drop (web-based software) Free http://slicedrop.com/
SPIERS Free www.spiers-software.org
tomviz Free http://www.tomviz.org/
VG Studio Max Commercial www.volumegraphics.com
Volocity Commercial http://www.improvision.com
VTK Free http://www.vtk.org/
KEKLIKOGLOU K. et al., Micro-computed tomography for natural history specimens
55
