Content uploaded by Daniil Nikitichev
Author content
All content in this area was uploaded by Daniil Nikitichev on Feb 07, 2018
Content may be subject to copyright.
3D printed ultrasound phantoms for clinical training
James Robertson,1* Emma Hill,1 Andrew A. Plumb,2 Simon Choong,3 Simeon J. West,3 Daniil I. Nikitichev1
1 Department of Medical Physics and Biomedical Engineering, University College London.
London, Gower Street, WC1E 6BT, London, United Kingdom.
2 Centre for Medical Imaging, University College London, London, United Kingdom
3 University College Hospital, 235 Euston Road, NW1 2BU, London, United Kingdom.
E-mail:
james.robertson.10@ucl.ac.uk and emma.hill.14@ucl.ac.uk
Internet: http://www.ucl.ac.uk/medphys
Ultrasound is a ubiquitous, portable structural imaging technique used to provide visual
feedback for a heterogeneous range of diagnostic and surgical techniques[1]. Preparatory training
for these techniques therefore demands a range of teaching models tailored for each application.
Existing anatomical models are often overly simple or prohibitively expensive, causing difficulties in
obtaining patient or procedure specific models. In this study we present ultrasound phantoms for
clinical teaching and training purposes, fabricated by extruding and “photopolymerization” three-
dimensional (3D) printing technologies.
3D printing is increasingly used in medicine, neuroscience and biotechnology due to its
affordability and the ability to create complex anatomical geometries[2]. 3D printed imaging
phantoms have been successfully demonstrated in the past but they were relatively simple[3][4]. In
order to create patient-specific models from medical image data, image-processing software is
required. In recent years, a number of image-processing methods have been developed that allow
the rapid generation of anatomically accurate 3D models from population or patient specific
data[5][6]. There is a growing interest in the application of these techniques in a clinical context for
the creation of anatomically accurate 3D printed models from medical images for therapeutic,
research and teaching applications [7-9].
For this study, clinical partners identified a requirement for practical rib and kidney imaging
phantoms for use in surgical teaching and planning of complex cases. We produced a series of
high quality 3D printed functional models of selected sets of ribs and a partial kidney model
suitable for use in clinical training for ultrasound guided therapy and diagnostics. Clinical partners
advised on specific design considerations for these models.
To produce these models, X-Ray CT data were segmented to extract volumes of interest. These
segmentations were transferred to MeshMixer software for refinement, before being exported as
stereolithography (STL) files for 3D printing software (Figure. 1.a). A rib models was printed in
enhanced PLA extruding material using an Ultimaker printer (Ultimaker, UK) while a kidney model
was fabricated in digital material (DM9860, Stratasys) using a polyjet Objet printer (Stratasys, UK)
(Figure 1.b,c). These models were combined in an ultrasound imaging phantom (Figure 1.c).
The ultrasonic profiles of the imagining phantom were obtained and compared with pre-existing
ultrasound images of the areas and organs of interest. These images were used to gauge
suitability of the phantom for teaching and surgical planning, and to determine the optimal 3D
printing material for use in realistic phantom production. Following review the final model design
will be of use to clinicians and educators for medical training purposes as the final STL files can be
easily shared between centres. Furthermore, the optimised production technique will be of interest
for anyone wishing to design a similar phantom for alternative medical applications.
Figure 1. a) Medial view of completed rib mesh in MeshMixer environment following segmentation
in Seg3D b) Completed 3D print of Rib model in Polylactic Acid (PLA) on Ultimaker 2 3D printer c)
Reduced scaled rib phantom with partial kidney phantom in fluid bath (insert: detail of kidney
phantom) d) Ultrasound scan of human rib
in-situ
, with reflection and shadowing from the rib
highlighted e) Ultrasound scan of imaging phantom showing shadowing and reflections from rib
and kidney phantoms.
Keywords
Ultrasound phantom, clinical training, image processing
Biography
James Robertson obtained an undergraduate degree in Biomedical Science with Medical Physics
from University College London in 2013 in part completion of an MBBS in Medicine (on hiatus). He
is currently reading for a PhD in Medical Physics at the Biomedical Ultrasound Group of the Dept.
of Medical Physics and Biomedical Engineering of UCL, under Dr Bradley Treeby. His research is
based on the use of numerical simulation of ultrasound for the transcranial focusing of ultrasonic
therapy, and includes the creation and validation of multiple ultrasonic bone phantoms.
References
[1] T. L. Szabo,
Elsevier Academic Press,
2004, 1-19.
[2] F. A. Zeiler, B. Unger, A. H. Kramer, A. W. Kirkpatrick, and L. M. Gillman,
Can. J. Neurol. Sci.
2013,
40,225.
[3] G. Sato dos Santos, E. Maneas, D. I. Nikitichev, A. Barburas, A. L. David, J. Deprest, A. Desjardins, T.
Vercauteren, S. Ourselin,
MICCAI 2015, Part I, Lect. Notes Comput. Sci.
(Navab, N., Hornegger, J.,
Wells, W. M. & Frangi, A. F.) 2015, 9349, 455.
[4] D. I. Nikitichev, A. Barburas, K. McPherson, S. J. West, J. M. Mari, and A. E. Desjardins,
Ultrasound
Med. Biol.
, 2015, in press.
[5] CIBC, "
http://seg3d.org"
2015.
[6] A. Fedorov, R. Beichel, J. Kalpathy-Cramer, J. Finet, J-C. Fillion-Robin, S. Pujol, C. Bauer, D. Jennings,
F. Fennessy, M. Sonka, J. Buatti, S. R. Aylward, J. V. Miller, S. Pieper, R. Kikinis,
Magn. Reson.
Imaging
2012, 30, 9.
[7] R. Partridge, N. Conlisk, J. A. Davies,
Organogenesis
2014, 8, 22.
[8] B. W. Turney,
J. Endourol.
2014, 28, 360.
[9] M. Kusaka, M. Sugimoto, N. Fukami, H. Sasaki, M. Takenaka, T. Anraku, T. Ito, T. Kenmochi, R.
Shiroki, K. Hoshinaga,
Transplant. Proc.
2015, 47, 596.