Conference PaperPDF Available

Tissue-equivalent phantoms recognition using hyperspectral imaging


Abstract and Figures

The incessant innovations of hyperspectral imaging (HSI) and data mining algorithms create necessity for developing reliable means of assessment and comparison. In medical applications of HSI, for instance, one of such means is tissue-equivalent phantoms. These phantoms are designed to mimic the spectral behavior of real living tissues. In this work, gel-based-phantoms were prepared with adjusted ingredients. The gel phantom's ingredients include India ink and Intralipid to provide absorption and scattering, respectively. Unlike visual assessment and photography, HSI was successful in identifying the various phantoms based on their spectral signature. In conclusion, we introduce a simple method of evaluating the performance of newly developed optical imaging techniques including HSI via affordable, inexpensive, and easy-to-make phantoms.
Content may be subject to copyright.
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
Military Technical College
Kobry El-Kobbah,
Cairo, Egypt
11th International Conference
on Electrical Engineering
Tissue-Equivalent Phantoms Recognition Employing Hyperspectral Imaging
Ramy Abdlaty* and Shirley Deng **
The incessant innovations, of hyperspectral imaging (HSI), and data mining algorithms,
express the necessity for developing reliable assessment and comparison means. In
medical applications of HSI, for instance, one of the means of assessment is tissue-
equivalent phantoms. These phantoms are designed to mimic the spectral behavior of
the real living tissues. In this work, gel-based-phantoms are prepared with altered
ingredients. The gel phantom’s ingredients include India ink, and Intralipid to provide
absorption and scattering respectively. Unlike visual assessment, and photography, HSI
for succeeded to identify the various phantoms based on its spectral signature. In
conclusion, we introduce a simple method to evaluate the performance of newly
developed optical imaging techniques including HSI via an affordable, inexpensive, and
easy to make phantoms.
1- Introduction
Until then, skin diseases diagnosis and treatment evaluation are inspected by visual
assessment, the gold standard [1], [2], via a dermatologist [3], [4]. However, visual
assessment was criticized in many studies to be subjective, qualitative, temporally
inconsistent, and invasive [5][10]. To overcome the critiques of visual assessment,
objective techniques were proposed for precise skin inspection/ treatment assessment.
The proposed objective techniques, in literature, include optical [11], and non-optical
techniques [12]. In optical techniques, there are two main approaches for skin
assessment; one of which is based on diffuse reflectance spectroscopic (DRS)
measurements while the other is based on color imaging. The spectroscopic-based-
approach is well-known for detecting the spectral signature of the skin’s symptoms by
high precision, inexpensive equipment [4]. Hence, it aids the dermatologist in
differentiating between visibly alike skin’s symptoms. Unfortunately, spectroscopic
measurements techniques need direct skin contact; which is hard to achieve in cases
including burns, and limited to a small area of inspection; therefore, it takes lengthy time,
and effort to judge a sizeable region of interest (ROI). Color imaging/ photography, unlike
spectroscopic measurements, is contactless and deals with a sizable ROI. Nonetheless,
* Egyptian Armed Forces.
** Faculty of Health sciences, McMaster University, Ontario, Canada.
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
photography suffers from poor spectral resolution; due to its limitation of three bands: red,
green and blue. Far from spectral shortage, photography is a cumbersome process since
it requires firm conditions in order to achieve the expected efficiency such as illumination
consistency. Taken together, the limitations of the famous optical techniques for skin
assessment provoked the inspiration of spectral imaging.
Spectral imaging (SI) is a hybrid technique combining both the advantages of DRS and
photography and thus, it overcomes the limitations of each [13]. HSI is used for acquiring
multiple frames for an object of interest, likewise photography, however, each frame is
captured at distinct wavelength [14]. As thus, the object’s ROI total spectral radiance is
acquired in a 3-dimensional data format, two of which are spatial and the third is spectral
one. SI data analysis/ classification enables a precise estimation of the componential
analysis for object under test [15]. Several considerations were taken in order to
categorize SI such as the number of bands, the resolution, and the acquisition schemes.
Regarding the number of bands, for example, SI is one of three: multispectral imaging
(MSI, usually uses less than 10 bands) [16][18], hyperspectral imaging (HSI, between
10 and 1000 bands) [13], [19], [20], or ultraspectral imaging (USI, 1000 or more bands)
[21]. In terms of the data acquisition, tunable filter-based-SI is common scheme for the
spectral scanning scheme. Two filters are most commonly used in HSI: acousto-optic
tunable filter (AOTF), and liquid crystal tunable filter (LCTF). AOTFs are distinguished by
higher spectral resolution, while LCTFs are recommended for achieving high quality
images. The use of tunable filters, in visible and near infrared spectra, entails the use of
broadband light sources such as tungsten-halogen or xenon light sources. Charge-
coupled device (CCD)/ complementary metal oxide semiconductor (CMOS) are the
challenging technologies for image detection in HSI. In terms of applications, HSI has a
successful role in several research fields including medicine [22]. Nevertheless, HSI
system is not handy for purchasing. Therefore, custom-made HSI systems is low-cost to
build but requires more effort to be validated and tested. To give an example, phantoms
with tissue like properties, provided that being longstanding, and easy to reproduce, are
fruitful way of optical innovative techniques assessment.
Tissue-equivalent phantoms (TEP) captivated the attention of research studies; since it
is convenient in simulating real tissues [23][27]. Typically, TEP fabrication is dependent
on wax, resin, and agarose as the base materials combined with either; milk, blood,
Intralipid, or yeast suspension. To mimic the natural absorbers/ diffusers in biologic
tissues, ink and polystyrene particles are added to the mixture [23], [25], [27][29]. TEP
can be customized to mimic the spectral signature of certain tissue in a minute band such
as red (633 nm) [30], [31] or a wider portion of the spectrum like near-infrared [32]. In this
work, we test a custom-made AOTF-HSI platform, originally designed by our group for
skin erythema assessment, by using visually alike, but physically different TEP. The test
results aid in evaluating the effectiveness of the developed AOTF-HSI system for a real
assessment regarding radiotherapy induced erythema in skin cancer treatment.
2- Materials and Methods
2.1- Hyperspectral Imaging Instrument
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
Conveniently, HSI setup uses a wide-band light source, to illuminate the object of interest.
The light source, used here, is a set of two commercial tungsten-halogen lamps with total
power of 300 watt. The reflected photons back from the illuminated ROI, within the field
of view (FOV) of the instrument, are collected and directed toward an optical filter
positioned ahead of an image detector to scan through the successive spectral bands of
operation. To give more details, Figure 1 displays the schematic diagram for the
developed AOTF-HSI optical setup in the current work. In the lead, there is a zoom lens
(Cannon EF-S 55-250 mm f/4-5.6 IS) installed in order to collect the reflected photons.
These photons, within the lens FOV, are focused to form the object image in its focal
plane. The aforesaid plane is adjusted to be at the focal plane for the second lens in the
HSI optical setup. The second lens is the leading lens of an optical relay, formed of two
typical achromatic lenses, acting as image transporter from the first focal plane to the
detector screen. The optical relay lenses are achromatic in order to avoid the chromatic
aberration. In between the relay lenses, a polarizing beam splitter (PBS) is installed. The
PBS function is to modify the randomly polarized input beam of photons into two
orthogonally polarized rays. One ray, of the two orthogonal rays, is transmitted through
the PBS while the second is perpendicularly reflected.
For high throughput purpose, a mirror followed by a half-wave plate (λ/2) are installed on
the reflecting side of the PBS. The reason for installing the mirror and the half-wave plate
is to modify the reflected ray in both direction and polarization. By this modification, the
two rays of the input beam of photons become not only polarized but also matching the
the AOTF polarization, and injected within its acceptance angle. The incorporated AOTF
(Gooch & Housego TF625-350-2-11-BR1A), here, is a noncollinear configuration. The
AOTF crystal is made of tellurium oxide crystal (TeO2 ) with an input aperture of 11 x 12
mm window. The filter is operating in the spectral range of 450-800 nm. An 8-channel
digital frequency synthesizer (Gooch & Housego MSD040-150-0.2ADM-A5H-8 x 1) drives
the AOTF device with a radio frequency (RF) signal ranges from 40-150 MHz The RF
driver is capable of simultaneously emitting 8 separate frequencies through the TeO2
crystal, by which 8 spectral bands are diffracted. A set of control commands are used to
adjust the tuning RF signal. These commands are sent through a universal serial bus
protocol via a work station. This work station controls not only the AOTF device, but also
the image detector. The image detector (Ximea MQ042rg-CM Enhanced-IR), employed
in the optical setup, is a monochromatic complementary metal oxide semiconductor
(CMOS) camera, with a sensor size of 11.24 mm2. The whole components are well fixed
to a metal platform to ensure the stability and rigidity of the HSI system during both
movement and operation.
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
Figure 1: A schematic diagram displays hyperspectral imaging (HSI) system
configuration used for AOTF and LCTF comparison. The optical components used to
build this setup are: zoom lens, square aperture, two achromatic (collimating and
focusing) lenses, flat mirrors; PBS: polarizing beam splitter; half-wave plate, and a
camera for capturing images, and computer for system control and data storage. All the
former components are tightly fixed to a metal platform to ensure the setup’s stability and
2.2-Tissue-equivalent phantom preparation
For reliability and marketing purposes, the newly developed optical instrumentation
including HSI need to be tested and characterized prior operation. One sort of
instrumentation testing is via the use of TEP’s. These phantoms are characterized by a
typical range of tissue’s optical properties. The easiest way of developing phantoms are
the aqueous suspensions, which are typically composed of Intralipid and India ink [24].
Although liquid phantoms are intensely adaptable, it is hard to include heterogeneities.
Therefore, several hydrogel-based phantoms mediated by agar, or gelatin offered the
opportunity to form distinct shapes and formats of phantoms.
Gel-based-phantoms, here, were constructed with identified concentrations of an
absorber (water soluble India ink) and scatterer (intralipid). The gel format of the
phantoms is controlled by using agar (R9012HR, Cedarlane Laboratories, Burlington, ON,
CA) solution by diluting agarose powder with a weight ratio of 1.0%. To verify the efficacy
of India-ink contributions used to prepare the phantom construction, samples of altered
ink solutions were prepared, displayed in Figure 2-(a), and characterized by two ways.
The first way is using a custom-made setup, shown in Figure 2-(b), composed of a broad
band light source, pinhole, objective lens, sample, and a fiber based spectrophotometer
(NEWPORT Model OSM400, Irvine, Ca, USA). The setup used a broad band light source
to illuminate the sample via a pinhole to clear the undesired optical noise. An objective
lens is placed at one focal length from the pinhole to to collimate the beam of light toward
the sample. The collimated beam, used to illuminate the ink samples is shaped using a
square aperture. The transmitted light is detected and measured via an optical fiber,
400 µm, and 0.22 NA, Thorlabs, Newton, NJ, USA) connected to a computer controlled
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
spectrophotometer. The second way of verification is using a high standard calibrated
spectrometer (GENYSISTM 10S UV-Vis, Thermo Fisher Scientific, Waltham, Ma, USA).
Dilutions of the absorber component were done using a stock solution of water soluble
India ink as shown in Figure 2. Toward constructing distinct phantoms, the level of optical
absorption value (mm-1) is changed in each phantom by increasing the ink incrementally
while holding the reduced scattering at a constant value.
Figure 2: (a) Dilutions of the India ink component, used in developing the gel-based
phantoms, were done using a stock solution of water soluble India ink. (b) A custom-made
optical setup for measuring the absorbance of varying solution of water soluble India ink.
In preparation process of TEP, the agar powder (0.2 gram) is added to 9.8 mL of distilled
water, and then heated in a water path (85 oC) for 50 minutes. In parallel, 8.6 mL of water,
1.2 mL of 20% stock Intralipid solution, and 0.2 mL of India ink are mixed and poured into
a 10 mL test tube. Once the agarose solution becomes transparent, the mixture
containing the Intralipid is also put in the water path for 10 minutes. After then, the two,
10 mL test tubes are combined in a 50 mL test tube to be well mixed. At this point, fast
work is critical because the gel will rapidly solidify once it is removed from the heated
water path. Finally, pour the 20 mL mixed solution into the 20 mL phantom to form the
shape shown in Figure 3. The whole process of phantom construction was carefully
considered in each stage to achieve the required reliability.
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
Figure 3: Tissue equivalent phantom (TEP’s) prepared with same ingredients (Agarose,
Intralipid, India ink, and distilled water) except the India ink concentrations. The ink
concentrations percentages are from left to right and from down to up are (0.0, 0.1, 0.2,
0.3 0.4, 0.5%) respectively.
2.3- Optical measurements
The optical measurement implemented, here, commenced with verifying the accuracy of
preparing samples of water soluble India ink. Five samples (0%,0.1%,0.2%,0.3%, and
0.4%) were used in the verification process. The measured light intensities before and
after inserting the samples were recorded. Based on Beer-Lambert’ s law, displayed in
Eq. (1), absorption was computed and plotted. Based on the former measurement, TEP,
with altered India ink concentration, were prepared. Afterwards, TEP were illuminated by
a wide band light source and imaged by the developed HSI device. An entire datacube
were recorded for each TEP. Instantaneously, each image cycle for a set of TEP’s, was
followed by capturing two datacubes. The first cube of images is taken for a white
standard reflectance target (Labsphere-SRS-99), while the second datacube is for the
sensor dark noise. The latter cube is recorded while the camera shutter and all the room
lights are turned off. The reflectance.
  
3- Experiments and Results
The prepared samples of ink solutions were characterized in terms of light transmittance
through samples. The results of the computed light intensity demonstrated an agreement
with the amounts of the India ink added to each sample. as shown in Figure 4. Based on
the attained results for sole India ink sole soluble solutions, the TEP’s were prepared in
the lab for multiple times. Other than India ink, the agarose and intralipid ratios were
adjusted to achieve the humidity and the stability of the phantoms. The stability of the
phantoms is of great importance in the imaging process since the employed HSI device
is, unlike microscope, horizontally imaging. The captured data is stored in the form of data
sets/ cubes for each phantom combined with the dark and the white standard target
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
Figure 4: the computed light absorption based on beer-Lambert’s law for the prepared
India ink soluble water solutions by (a) custom-made spectrometric measurement, and
(b) high standard calibrated spectrometer (GENYSISTM 10S UV-Vis, Thermo Fisher
Scientific, Waltham, Ma, USA).
The TEP’s datacubes were processed: first, by detector noise removal via subtracting the
dark cubes. Afterwards, the TEP’s datacubes are normalized to the formerly mentioned
white standard reflectance target datacube, in order to accommodate for uneven
illumination effects and the spectral dependence of the imaging setup applying Eq. (2).
Where is the normalized reflectance,  is the measured back reflected intensity of
the white standard, and  is the reflectance of the TEP. The value for is plotted
against one wavelength (~ 700 nm) to show how the reflectance of the prepared
phantoms are interrelated together as shown in Figure 5.
  
  
350 400 450 500 550 600 650 700 750 800 850
Wavelength (nm)
0.1% Ink
0.2% Ink
0.3% Ink
0.4% Ink
350 400 450 500 550 600 650 700 750 800 850
Wavelength (nm)
0.1% ink
0.2% ink
0.3% ink
0.4% ink
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
Figure 5: Computed TEP’s reflectance at the spectral band of the centered 696 nm
wavelength for 5 phantoms with gradual increase in India ink ingredients. A linear
trendline is drawn to display the error between the computed reflectance and the ideal
4- Discussion
Past, current, and continued efforts are exerted to compose phantoms in order to mimic
human tissues’ behavior. These efforts are recorded in many studies in literature [33].
Literally, the aqueous solution of a scattering medium, including Intralipid, and a water-
soluble India ink is designated as one of the top popular sort of biological simulating
phantoms. However, this sort of phantoms suffers from the disadvantage of the
incapability of creating layered structures. To overcome the previously mentioned
disadvantage, agarose component is involved in synthesizing phantoms. The agar
component allowed the development of multilayered structures where absorbing and
scattering ingredients are included. Likewise, the phantoms prepared for this work can be
molded into any custom shape.
The work, implemented here, made use of the agar-based phantoms to assess the
sensitivity of HSI in detecting changes in embeded chromophores concentration within
the visible and the near infrared region (450:800nm). The variation of the ink dye
embeded in the phantoms, used here, mimics the process of the human skin optical
properties change due to the variation of its encompassed chromophores. To verify the
reliability of the optical properties of the phantoms, the preparation process was repeated
and revised provided that the absorbing and scattering components are accurately
speckled. The revised steps of preparation include checking the macroscopic
homogeneity of the phantom ingredients by measuring the transmittance as has been
done with the India ink in Figure 4.
The procedure we introduce, in this work, facilitates the development of a phantom which
is straightforward in manufacturing, usage, and molding in alternate formats. The former
advantages of the phantom make it useful in biomedical studies, for example, offering the
required in vivo environment, likewise contrast the light-transport models with empirical
measurements. To give this work more sense, the phantoms, used here, proved the
feasibility of employing HSI in a research study, currently in its way at McMaster
university, in order to diagnose the skin lesions treated via radiotherapy. Figure 5
Ph-0.1 Ph-0.2 Ph-0.3 Ph-0.4 Ph-0.5
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
illustrates the capability of HSI for differentiating between various TEP while visual
assessment is incapable.
5- Conclusion
In a nutshell, the study presented here in this paper tagged the AOTF-HSI, with the
capability of differentiating between TEP’s with dissimilar ingredients. As a consequence,
the decision of using HSI becomes more feasible in skin diseases diagnosis. The major
advantages of AOTF is summarized in the following points:
AOTF-HSI configuration is distinguished by:
a. Fine spectral resolution
b. Significant out of band suppression
c. Fast wavelength access, appropriate for video rate
d. High diffraction efficiency (90%)
e. Compact size
f. Simultaneous diffraction of 8-distinct wavelengths
Ultimately, we need to emphasize that this work is a preliminary step toward skin
erythema objective assessment. In general, HSI system robust design is application
based. For medical applications, as an example, if the biological tissue of interest has
unknown optical properties, or characterized by close spectral features, AOTF-HSI might
be the recommended option.
The ingredients of the tissue equivalent phantoms used in the experiments were kindly
provided by Professor Qiyin Fang and described by Sharon Goh.
[1] A. S. C. J. Taylor S.C., “The taylor hyperpigmentation scale: A new visual
assessment tool for the evaluation of skin color and pigmentation,” Ther. Clin.,
vol. 76, no. 4, pp. 270274, 2005.
[2] D. Basketter, F. Reynolds, M. Rowson, C. Talbot, and E. Whittle, “Visual
assessment of human skin irritation: a sensitive and reproducible tool.,” Contact
Dermatitis, vol. 37, pp. 218220, 1997.
[3] S. G. Kong, M. Martin, and T. Vo-Dinh, “Hyperspectral Fluorescence Imaging for
Mouse Skin Tumor Detection,” ETRI J., vol. 28, no. 6, pp. 770776, Dec. 2006.
[4] D. L. Glennie, J. E. Hayward, D. E. Mckee, and T. J. Farrell, “Inexpensive diffuse
reflectance spectroscopy system for measuring changes in tissue optical
properties,” J. Biomed. Opt., vol. 19, no. December, pp. 105005-1-1050056,
[5] A. D. Ormerod, C. M. Dwyer, R. Weller, D. H. Cox, and R. Price, “A comparison of
subjective and objective measures of reduction of psoriasis with the use of
ultrasound, reflectance colorimetry, computerized video image analysis, and nitric
oxide production,” J. Am. Acad. Dermatol., vol. 37, no. 1, pp. 5157, 1997.
[6] J. Nyström, P. Geladi, B. Lindholm-Sethson, J. Rattfelt, A. C. Svensk, and L.
Franzen, “Objective measurements of radiotherapy-induced erythema,” Ski. Res.
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
Technol., vol. 10, no. 4, pp. 242250, 2004.
[7] G. V Oliveira, D. Chinkes, C. Mitchell, G. Oliveras, H. K. Hawkins, and D. N.
Herndon, “Objective assessment of burn scar vascularity, erythema, pliability,
thickness, and planimetry.,” Dermatol. Surg., vol. 31, pp. 4858, 2005.
[8] A. Raina, R. Hennessy, M. Rains, J. Allred, D. Diven, and M. K. Markey,
“Objective measurement of erythema in psoriasis using digital color photography
with color calibration,” Ski. Res. Technol., no. 7, pp. 16, 2015.
[9] L. Diana L. Glennie, Joseph E. Hayward, Orest Z. Ostapiak, James Wright and T.
J. F. Doerwald-Munoz, “Diffuse reflectance spectroscopy for monitoring erythema
in head & neck intensity modulated radiation therapy,” J. Radiat. Oncol., 2014.
[10] A. Madooei et al., “Hyperspectral Image Processing for Detection and Grading of
Skin Erythema,” in SPIE_MI.
[11] N. Kollias and G. N. Stamatas, “Optical non-invasive approaches to diagnosis of
skin diseases,” J. Investig. Dermatology Symp. Proc., vol. 7, pp. 6475, 2002.
[12] B. Jung, S. Kim, Y. Bae, H. Kang, Y. Lee, and J. S. Nelson, “Real-time
measurement of skin erythema variation by negative compression: pilot study.,” J.
Biomed. Opt., vol. 17, no. 8, pp. 814221, Aug. 2012.
[13] K. B. Yushkov and V. Y. Molchanov, “Hyperspectral imaging acousto-optic system
with spatial filtering for optical phase visualization,” J. Biomed. Opt., vol. 22, no. 6,
p. 66017, 2017.
[14] R. E. Warren and D. B. Cohn, “Chemical detection on surfaces by hyperspectral
imaging,” J. Appl. Remote Sens., vol. 11, no. 1, p. 15013, 2017.
[15] G. Lu, D. Wang, and Z. G. Chen, “Spectral-spatial classification for noninvasive
cancer detection using hyperspectral imaging detection using hyperspectral
imaging,” J. Biomed. Opt., vol. 19(10), 2014.
[16] L. Kong et al., “Handheld erythema and bruise detector,” Proc. SPIE, vol. 6915, p.
69153K69153K7, 2008.
[17] D. Jakovels, J. Spigulis, and I. Saknite, “Multi-spectral mapping of in vivo skin
hemoglobin and melanin,” SPIE Proc., vol. 7715, p. 77152Z--77152Z, 2010.
[18] L. Kong et al., “Single sensor that outputs narrowband multispectral images.,” J.
Biomed. Opt., vol. 15, no. 1, p. 10502, 2010.
[19] S. Miclos, S. V. Parasca, M. A. Calin, D. Savastru, and D. Manea, “Algorithm for
mapping cutaneous tissue oxygen concentration using hyperspectral imaging,”
Biomed. Opt. Express, vol. 6, no. 9, pp. 34203430, 2015.
[20] B. S. Sorg, B. J. Moeller, O. Donovan, Y. Cao, and M. W. Dewhirst,
“Hyperspectral imaging of hemoglobin saturation in tumor microvasculature and
tumor hypoxia development.,” J. Biomed. Opt., vol. 10, no. 4, p. 44004, 2014.
[21] a. D. Meigs, I. Otten, L.J., and T. Y. Cherezova, “Ultraspectral imaging: a new
contribution to global virtual presence,” in 1998 IEEE Aerospace Conference
Proceedings (Cat. No.98TH8339), 1998, vol. 2, no. October.
[22] G. Lu and B. Fei, “Medical hyperspectral imaging: a review.,” J. Biomed. Opt., vol.
19, no. 1, p. 10901, Jan. 2014.
[23] J. C. Ramirez-san-juan, B. Choi, W. Franco, J. S. Nelson, and G. Aguilar, “Effect
of ambient humidity on light transmittance through skin phantoms during cryogen
spray cooling,” Phys. Med. Biol., vol. 50, pp. 18, 2005.
[24] J. Pekar and M. S. Patterson, “Fabrication and characterization of phantoms with
Proceedings of the 11th ICEENG Conference, 3-5 April, 2018
tissue-like optical properties from 500 to 700 nm,” Med. Laser Appl., vol. 25, pp.
147153, 2010.
[25] N. Chahat, M. Zhadobov, S. Alekseev, and R. Sauleau, “Human skin-equivalent
phantom for on-body antenna measurements in 60 GHz band,” Electron. Lett.,
vol. 48, no. 2, pp. 910, 2012.
[26] J. R. Weber et al., “Multispectral imaging of tissue absorption and scattering using
spatial frequency domain imaging and a computed-tomography imaging
spectrometer.,” J. Biomed. Opt., vol. 16, no. 1, p. 11015, 2011.
[27] I. Nishidate, T. Maeda, K. Niizeki, and Y. Aizu, “Estimation of melanin and
hemoglobin using spectral reflectance images reconstructed from a digital RGB
image by the Wiener estimation method,” Sensors (Switzerland), vol. 13, no. 6,
pp. 79027915, 2013.
[28] J. W. Pickering, S. a Prahl, N. van Wieringen, J. F. Beek, H. J. Sterenborg, and
M. J. van Gemert, “Double-integrating-sphere system for measuring the optical
properties of tissue.,” Appl. Opt., vol. 32, no. 4, pp. 399410, 1993.
[29] S. C. Gebhart, R. C. Thompson, and A. Mahadevan-Jansen, “Liquid-crystal
tunable filter spectral imaging for brain tumor demarcation,” Appl. Opt., vol. 46,
no. 10, p. 1896, 2007.
[30] M. S. Patterson, B. Chance, and B. C. Wilson, “Time resolved reflectance and
transmittance for the non-invasive measurement of tissue optical properties,”
Appl. Opt., vol. 28, no. 12, pp. 23312336, 1989.
STERENBORG, “Skin Optics,” IEEE Trans. Biomed. Eng., vol. 36, no. 12, pp.
11461154, 1989.
[32] M. Firbankt, M. Odq, and D. T. Delpyt, “An improved design for a stable and
reproducible phantom material for use in near-infrared spectroscopy and
imaging,” Physc. Med. Biol., vol. 40, pp. 955961, 1995.
[33] M. Lualdi, A. Colombo, B. Farina, and S. Tomatis, “A Phantom With Tissue-Like
Optical Properties in the Visible and Near Infrared for Use in Photomedicine,”
Lasers Surg. Med., vol. 28, no. July 2000, pp. 237243, 2001.
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
We have developed a method for using hyperspectral (HS) data to identify and locate chemical materials on arbitrary surfaces using the materials' reflection or emission spectra that makes no prior modeling assumptions about the presence of pure pixels or the statistics of the background clutter and sensor noise. To our knowledge, this is the first time that surface detection without dependence on background information has been achieved. There are three main components to the method: (1) an HS unmixing algorithm based on the alternating direction method of multipliers that is applied over local subsets of the imaging to resolve the HS data into a set of linearly independent spectral and spatial components; (2) the fitting of those unmixing spectra to a set of candidate template spectra; and (3) a support vector machine classifier for chemical detection, identification, and location. The algorithm is illustrated on HS data collected by a Telops Hyper-Cam infrared camera on data resulting from the deposition of chemical agent simulants on various surfaces.
Full-text available
The measurement of changes in blood volume in tissue is important for monitoring the effects of a wide range of therapeutic interventions, from radiation therapy to skin-flap transplants. Many systems available for purchase are either expensive or difficult to use, limiting their utility in the clinical setting. A low-cost system, capable of measuring changes in tissue blood volume via diffuse reflectance spectroscopy is presented. The system consists of an integrating sphere coupled via optical fibers to a broadband light source and a spectrometer. Validation data are presented to illustrate the accuracy and reproducibility of the system. The validity and utility of this in vivo system were demonstrated in a skin blanching/reddening experiment using epinephrine and lidocaine, and in a study measuring the severity of radiation-induced erythema during radiation therapy. (C) 2014 Society of Photo-Optical Instrumentation Engineers (SPIE)
Full-text available
Early detection of malignant lesions could improve both survival and quality of life of cancer patients. Hyperspectral imaging (HSI) has emerged as a powerful tool for noninvasive cancer detection and diagnosis, with the advantage of avoiding tissue biopsy and providing diagnostic signatures without the need of a contrast agent in real time. We developed a spectral-spatial classification method to distinguish cancer from normal tissue on hyperspectral images. We acquire hyperspectral reflectance images from 450 to 900 nm with a 2-nm increment from tumor-bearing mice. In our animal experiments, the HSI and classification method achieved a sensitivity of 93.7% and a specificity of 91.3%. The preliminary study demonstrated that HSI has the potential to be applied in vivo for noninvasive detection of tumors. (C) 2014 Society of Photo-Optical Instrumentation Engineers (SPIE)
Full-text available
Hyperspectral imaging (HSI) is an emerging imaging modality for medical applications, especially in disease diagnosis and image-guided surgery. HSI acquires a three-dimensional dataset called hypercube, with two spatial dimensions and one spectral dimension. Spatially resolved spectral imaging obtained by HSI provides diagnostic information about the tissue physiology, morphology, and composition. This review paper presents an overview of the literature on medical hyperspectral imaging technology and its applications. The aim of the survey is threefold: an introduction for those new to the field, an overview for those working in the field, and a reference for those searching for literature on a specific application.
We report a method for phase visualization in the images of transparent specimens using analog image processing in incoherent light. The experimental technique is based on adaptive bandpass spatial filtering with an amplitude mask matched with an acousto-optic tunable filter in a telecentric optical system. We demonstrate the processing of microscopic images of unstained and stained histological sections of human thyroid tumor with improved contrast. © 2017 Society of Photo-Optical Instrumentation Engineers (SPIE).
The measurement of tissue oxygenation plays an important role in the diagnosis and therapeutic assessment of a large variety of diseases. Many different methods have been developed and are currently applied in clinical practice for the measurement of tissue oxygenation. Unfortunately, each of these methods has its own limitations. In this paper we proposed the use of hyperspectral imaging as new method for the assessment of the tissue oxygenation level. To extract this information from hyperspectral images a new algorithm for mapping cutaneous tissue oxygen concentration was developed. This algorithm takes into account and solves some problems related to setting and calculation of some parameters derived from hyperspectral images. The algorithm was tested with good results on synthetic images and then validated on the fingers of a hand with different blood irrigation states. The results obtained have proved the ability of hyperspectral imaging together with the developed algorithm to map the oxy- and deoxyhemoglobin distribution on the analyzed fingers. These are only preliminary results and other studies should be done before this approach to be used in the clinical setting for the diagnosis and monitoring of various diseases.
Traditional metrics for evaluating the severity of psoriasis are highly subjective, which complicates efforts to identify effective treatments in clinical trials. We propose a method for the objective measurement of the psoriasis severity parameter of erythema (redness). This procedure is standardized for different camera systems and lighting environments through the usage of a color card with predetermined color values in order to calibrate the images. Quantitative measures based on the digital color images are shown to correlate well with subjective assessment of psoriasis severity collected using a standard numerical scale by a panel of dermatologists. Additionally, the color calibration process is shown to improve results.
The multi-spectral imaging technique has been used for distant mapping of in-vivo skin chromophores by analyzing spectral data at each reflected image pixel and constructing 2-D maps of the relative concentrations of oxy-/deoxyhemoglobin and melanin. Instead of using a broad visible-NIR spectral range, this study focuses on narrowed spectral band 500-700 nm, so speeding-up the signal processing procedure. Regression analysis confirmed that superposition of three Gaussians is optimal analytic approximation for the oxy-hemoglobin absorption tabular spectrum in this spectral band, while superposition of two Gaussians fits well for deoxy-hemoglobin absorption and exponential function - for melanin absorption. The proposed approach was clinically tested for three types of in-vivo skin provocations - ultraviolet irradiance, chemical reaction with vinegar essence and finger arterial occlusion. Spectral range 500-700 nm provided better sensitivity to oxy-hemoglobin changes and higher response stability to melanin than two reduced ranges 500-600 nm and 530-620 nm.
Visual inspection of intact skin is commonly used when assessing persons for pressure ulcers and bruises. Melanin masks skin discoloration hindering visual inspection in people with darkly pigmented skin. The objective of the project is to develop a point of care technology capable of detecting erythema and bruises in persons with darkly pigmented skin. Two significant hardware components, a color filter array and illumination system have been developed and tested. The color filter array targets four defined wavelengths and has been designed to fit onto a CMOS sensor. The crafting process generates a multilayer film on a glass substrate using vacuum ion beam splitter and lithographic techniques. The illumination system is based upon LEDs and targets these same pre-defined wavelengths. Together, these components are being used to create a small, handheld multispectral imaging device. Compared to other multi spectral technologies (multi prisms, optical-acoustic crystal and others), the design provides simple, low cost instrumentation that has many potential multi spectral imaging applications which require a handheld detector.