Water-sensitive Gelatin Phantoms for Skin Water Content Imaging

Gennadi Saiko

and Alexandre Douplik

Swift Medical Inc, 1 Richmond St. W., Toronto, Canada

Department of Physics, Ryerson University, Toronto, Canada

Keywords: Multispectral Imaging, Tissue Phantoms, Skin Water Content, Skin Moisture.

Abstract: Oxygen supply to tissues can be seriously impacted during wound healing. Edema (accumulation of fluids in

interstitial space) can increase the distance between capillaries, thus decreasing oxygen supply to cells. There

is no standard clinical tool for quantification of edema, and early edema detection (preferably preclinical) is

of great clinical need. Multispectral imaging can be a helpful clinical tool to characterize water content in the

skin. However, to develop and validate this technology, a reliable water-sensitive preclinical model has to be

developed. The scope of this work is to develop a water-responsive skin model and assess the feasibility of

extracting water content using multispectral imaging. Methods: A phantom fabrication protocol has been

developed. The phantoms are based on the gelatin crosslinked with glutaraldehyde. TiO2 nanoparticles were

added to mimic the optical properties of the skin. To emulate various water content, phantoms were dipped

in the water for various duration. The phantoms were imaged using the Multi-Spectral Imaging Device

(MSID) (Swift Medical Inc, Toronto). MSID is a multispectral imaging system for visualization of tissue

chromophores in surface tissues. It uses 12-bit scientific-grade NIR-enhanced monochrome camera (Basler,

Germany) and ten wavelength light source (600-1000nm range) to visualize the distribution of oxy-,

deoxyhemoglobins, methemoglobin, water, and melanin. The imaging distance is 30cm, the field of view:

7x7cm. Results: Initial results show that the developed model mimics the optical scattering properties of the

skin. MSID was able to extract water content using a full set (ten wavelengths) and a subset (three

wavelengths) of channels. Conclusions: A new water responsive model for skin moisture imaging has been

developed. Initial experiments with multispectral imaging of these phantoms show feasibility of tissue water

content imaging with Si-based cameras.

1 INTRODUCTION

Edema (accumulation of fluids in interstitial space) is

a common clinical sign in a wide variety of

conditions. Being a nonspecific finding it often poses

a challenge for the clinician. While in many cases, it

has a benign origin, in other instances, it can be a sign

of life-threatening conditions. Because the

interstitium can easily accommodate several liters of

fluid, a patient’s weight may increase by nearly 10%

before pitting edema is evident. Thus, early detection

of edema (preferably preclinical) is of great

importance, especially for patients with compromised

health (diabetes, kidney or heart conditions, etc.).

Edema can seriously impact wound healing by

restricting oxygen supply to tissues. For example,

https://orcid.org/ 0000-0002-5697-7609

https://orcid.org/ 0000-0001-9948-9472

edema can increase the distance between capillaries,

thus decreasing oxygen supply to cells, or it may

compress small vessels to shut off the local blood

supply at all, thus creating necrotic tissue.

There is no standard for an objective measurement

of edema. In particular, for peripheral edema, the

most widely-used technique is a subjective clinical

assessment where an examiner applies pressure with

an index finger to the patient’s ankle (Seidel, 1995) to

capture pit depth and the time needed for the skin to

return to its original state (recovery time). Despite

common use, this method has not been proven to be a

sufficiently objective, reliable, or sensitive

assessment of edema. Several quantitative methods to

measure peripheral edema have been proposed, but

they are mostly used in physical therapy and sports

medicine: patient questionnaire, ankle circumference

130

Saiko, G. and Douplik, A.

Water-sensitive Gelatin Phantoms for Skin Water Content Imaging.

DOI: 10.5220/0008919501300134

In Proceedings of the 13th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2020) - Volume 2: BIOIMAGING, pages 130-134

ISBN: 978-989-758-398-8; ISSN: 2184-4305

(Mora, 2002), figure-of-eight (Mawdsley, 2000),

edema tester (Cesarone, 1999), indirect leg volume

(by series of ankle/leg circumferences) (Latchford,

1997), and water-displacement volumetry (Kaulesar

Sukul, 1993).

Several quantitative technologies have been used

to assess edema or the hydration of skin non-

invasively. They include skin impedance methods,

ultrasound, and magnetic resonance imaging, and

spectroscopy.

The major part of existing methods of measuring

of edema suffers from various shortcomings; namely,

subjectivity (Brodovicz, 2009), inability to detect on

early stages (Hedlund, 1985), and impossibility to be

applied in certain clinical or field settings (e.g., water

displacement for postoperative patients, MRI, and

ultrasound in field settings).

Thus, it would be useful to have a more widely

accessible way to investigate edema and ideally

visualize it in various clinical and field settings.

Optical spectroscopy and imaging, along with the

skin impedance measurements (Mayrovitz, 2007), are

promising modalities for the non-invasive diagnosis

and monitoring of skin water content. However,

optical imaging has several inherited advantages over

skin impedance measurements: a) it is able to

visualize (image) the large field of view, while skin

impedance techniques are essentially one point

measurements, prone to operator’s errors, b) it is a

remote measurement, while skin impedance requires

contact with the skin and sterilization after each use.

Thus, optical water content imaging would be a

useful clinical tool adjuvant to diagnostics of

peripheral vascular disease and pressure injuries.

However, the development of such imaging

technology is complicated in part due to the lack of

the gold standard and established water responsive

models. For example, its comparison and/or

validation with skin impedance techniques is not a

straightforward task such as various geometries of

skin impedance probes have different sampling

depths, which can be incomparable with an optical

sampling depth at a particular wavelength.

Thus, one of the important steps in developing the

water content imaging modality is to develop a

controllable water-responsive model, which can be

used to validate/calibrate the technology in the

absence of the gold standard.

While multiple experimental (Pogue, 2006;

Ohmae, 2018) and computational (Kainz, 2018)

models have been developed to mimic the optical and

RF properties of the human tissues, still there is a

need in versatile water-responsive models.

Our group works on multispectral imaging moda-

lities adjuvant to tissue oximetry. In our previous

works, we have developed several phantom models

and demonstrated the feasibility of multispectral

visualization of methemoglobin in the tissue (Saiko,

2018).

This scope of this work is to develop a water-

responsive skin model and assess the feasibility of

multispectral imaging of water content in the skin.

2 METHODS

Water content within the stratum corneum gradually

increases from about 10% to about 30% between the

surface and deeper layers, followed by an abrupt

increase to about 70% in the epidermis (Warner,

1988).

The water content of the different skin layers is of

interest in different fields. While the water content of

the epidermis is the primary interest in the skincare

industry, subepidermal water accumulation is of

interest in wound care.

Water content imaging of the different skin layers

can be based on various light absorption bands. Water

absorption at 1440nm is 30 times stronger than at

1190nm, which in turn more than two times stronger

than absorption at 970nm. Given this hierarchy and

corresponding light penetration depths it is possible

to expect that 1440nm and 1920nm wavelengths are

suitable for imaging of water content in uppermost

skin layers (stratum corneum), while 970nm

and1190nm can be used for water content

determination and imaging in deeper skin layers,

including epidermis, dermis (1190nm) and even

subcutaneous tissues (970nm). Illumination of the

skin with these wavelengths will integrate signals

from various depths (up to a few millimeters) and thus

will not be sensitive to skin conditions in surface

layers, which due to their small thickness and low

water content account only for a small fraction of

water in integration volume.

For our purposes, we selected 970nm range,

which a) have the required sampling depth (dermis

and subcutaneous tissues), and b) can be implemented

using inexpensive Si-based sensors.

To explore the feasibility of quantification of

water content in the skin, we have developed a water-

responsive skin model. The model is based on a

mechanical gelatin-based human skin model

(Dabrowska, 2017) with adaptations to mimic the

optical parameters of the skin.

Water-sensitive Gelatin Phantoms for Skin Water Content Imaging

131

2.1 Materials

The phantoms were based on the type A gelatine

derived from porcine skin, with 300 g gel strength

(G2500, Sigma-Aldrich, Canada), which does not

have significant optical absorption and scattering in

the visible range. Each phantom contained 10%

(w/w) of type A gelatin.

TiO2 particles (Sigma-Aldrich, Canada) were

used to mimic the scattering property of the skin.

Glutaraldehyde (Sigma Aldrich, Canada) was

used for the cross-linking of gelatin.

2.2 Phantom Fabrication

We have developed the following protocol for

phantom fabrication:

1. Prepare a 10 wt% solution of gelatine (type A,

bloom no 300, Sigma Aldrich) in distilled water

by continuous stirring at 60 °C for 2 h.

2. Mix with a known concentration of a scattering

agent (TiO2 nanoparticles).

3. Place the final solution in an ultrasonic bath at a

temperature of 37 – 38 degrees to degas for 2

minutes to disrupt the air bubbles trapped inside

the phantom,

4. Pour the solution into the Petri dish (with wax

paper layer) to solidify: 2mm layer.

5. Leave the phantom to dry for 24 h at room

temperature.

6. Gelatin crosslinking: Place the phantom in 1 wt%

solution of glutaraldehyde (Sigma Aldrich) in

Dulbecco's PBS buffer (DPBS, GIBCO) for 24 h

at room temperature under continuous stirring

(130 rpm).

7. Rinse phantoms with distilled water.

8. Dry phantoms by wrapping in paper towels and

placing between two boards with the use of

weight, to avoid ripples caused by drying-related

contraction.

9. Change paper towels daily and measure the

weight of each phantom. It is considered to be dry

after mass stabilization (about 6 days).

To emulate various hydration levels in the skin, the

samples were dipped in the water for various time.

Electronic scales measured weight before and after

dipping.

2.3 Imaging

The Multi-Spectral Imaging Device (MSID) (Swift

Medical Inc, Toronto) is a multispectral imaging

system (see Figure 1) for visualization of the

distribution of oxy-, deoxyhemoglobins,

methemoglobin, water, and melanin in the skin. The

MSID consists of a 10-channel illumination unit, a

scientific-grade camera (a 12-bit NIR-enhanced

monochrome camera acA1300-60gmNIR (Basler,

Germany)), and a processing unit, which coordinates

them and collect data. Ten channels of the

illumination unit illuminate the target area with 630,

660, 690, 735, 810, 830, 850, 880, 940, and 970nm,

respectively. Each channel consists of 4 high power

LEDs arranged into a circle. The illumination unit

produces a sequence of light flashes, each flash at a

particular wavelength, while the camera captures a

series of images, each with illumination at particular

wavelength. The acquired images are arranged into a

3D hypercube (, i, j) for further processing. The

imaging distance is 30cm, the field of view: 7x7cm.

2.4 Image Processing

During each measurement, the MSID device captures

11 images: ten with illumination at a particular

wavelength and one without additional illumination

(ambient light only). The processing consists of the

following consequential steps: a) calculate diff

images (subtract the image without additional

illumination from the image with illumination at a

particular wavelength), b) obtain reflectance images

by dividing the diff image on the diff image of the

reference object, c) extract index of absorption 

from reflectance using tissue light transport model

(e.g., Beer-Lambert), d) extract tissue chromophore

concentrations using least square fitting.

To emulate a compact device scenario, various

subsets of captured images (10 or 3 of them) were

used to extract water content.

The MSID device uses the 12bit scientific-grade

camera (acA1300-60gmNIR (Basler, Germany)).

False-color maps were used to visualize water

content.

Figure 1: Multi-Spectral Imaging Device (MSID).

BIOIMAGING 2020 - 7th International Conference on Bioimaging

132

3 RESULTS

Rough calculations based on the model developed in

(Saiko, 2019) show that we can expect approximately

-0.4% change in reflectance for every 1% increase in

the absorption coefficient at 970nm. Given that the

water is the primary absorber in this range we can

expect that it will be possible to detect even

preclinical peripheral edema with an expected

increase in water content. In particular, we expect that

a 50% increase in interstitial fluid volume for water

content in the skin, adipose tissue, connective tissue

will translate into a 5-6% increase in total tissue water

content (see Table 1). This increase will translate into

-2.0...-2.4% change in tissue reflectance, which is

possible to detect even using an 8-bit camera.

The developed phantoms were visualized using

MSID. To emulate various water content in the tissue,

phantoms were dipped in the water for the various

duration.

Table 1: Initial and expected water content for different

tissue types.

Tissue

Initial water

content, %

(Braunwald,

1994)

Expected water

content for 50%

increase in IFV, %

Skin 72 79

Skeletal muscle 76 77.4

Adipose tissue 14 18.5

Connective tissue 80 86

Figure 2: Water content extraction using ten wavelengths.

Dry sample (upper left), sample dipped for several seconds

(top right), 5 min (bottom left), and 20 min (bottom right).

The results of water content extraction using all

ten channels and three channels are presented in

Figure 2 and Figure 3, respectively. Four samples on

these figures have various water content: dry sample

(upper left), and samples dipped for several seconds

(top right), 5 min (bottom left) and 20 min (bottom

right).

Figure 3: Water content extraction using three wavelengths.

Dry sample (upper left), sample dipped for several seconds

(top right), 5 min (bottom left), and 20 min (bottom right).

4 DISCUSSION

Our initial calculations show that the

multispectral/hyperspectral imaging using a 970nm

band and Si-based sensor can retrieve clinically

relevant changes in the tissue water content.

These calculations were supported by our initial

experiments on phantoms, which demonstrate the

feasibility of water content imaging using Si-based

cameras. The water content was successfully

extracted using the full set of channels (10 channels)

and its subset (3 channels, including 970nm). Results

are almost identical, so we can conclude that three

illumination wavelengths may be sufficient for water

content visualization. These results are in line with

preliminary experiments on volunteers reported

separately (Saiko, 2020).

Initial experiments also show that the developed

gelatine-based phantoms a) demonstrate optical

properties similar to the skin, and b) demonstrate

required water-responsive properties. We will report

phantom validation results separately. The current

study with dipping phantoms in the water was a

proof-of-concept. In the future, we plan to vary

phantom water content in a more quantifiable way by

spreading a certain amount of water over its surface

and comparing water content results with those

obtained by bioimpedance measurements.

It should be noted that the cross-linked phantoms

changed their color from clear to pink during

crosslinking. In future work, we plan to investigate

Water-sensitive Gelatin Phantoms for Skin Water Content Imaging

133

these color changes and find conditions that minimize

them or investigate other cross-linking agents.

5 CONCLUSIONS

A new water responsive model for skin water content

imaging has been developed. Initial experiments with

multispectral imaging of these phantoms show the

feasibility of tissue water content imaging with Si-

based cameras using a 970nm band.

ACKNOWLEDGEMENTS

Authors are thankful to Burhan Hussein and Andrei

Betlen for help with phantom fabrication and image

processing.

REFERENCES

Seidel H.M, Ball J.W, Dains J.E, et al., 1995. Heart and

blood vessels. In: Schrefer S, ed. Mosby’s Guide to

Physical Examination, St. Louis, MO: Mosby. 3rd ed.

Mora S, Zalavras C.G, Wang L, Thordarson DB, 2002. The

role of pulsatile cold compression in edema resolution

following ankle fractures: a randomized clinical trial.

Foot Ankle Int, 23:999-1002.

Mawdsley R.H, Hoy D.K, Erwin P.M, 2000. Criterion-

related validity of the figure-of-eight method of

measuring ankle edema. J Orthop Sports Phys Ther,

30:149-153.

Cesarone M.R, Belcaro G, Nicolaides A.N, Arkans E,

Laurora G, De Sanctis MT, Incandela L, 1999. The

edema tester in the evaluation of swollen limbs in

venous and lymphatic disease. Panminerva Med;41:10-

14.

Latchford S, Casley-Smith J.R, 1997. Estimating limb

volumes and alterations in peripheral edema from

circumferences measured at different intervals.

Lymphology; 30:161-164.

Kaulesar Sukul D.M, den Hoed P.T, Johannes E.J, van

Dolder R, Benda E., 1993. Direct and indirect methods

for the quantification of leg volume: comparison

between water displacement volumetry, the disk model

method and the frustum sign model method, using the

correlation coefficient and the limits of agreement. J

Biomed Eng; 15:477-480.

Brodovicz K.G, McNaughton K, Uemura N, Meininger G,

Girman CJ, Yale SH., 2009. Reliability and Feasibility

of Methods to Quantitatively Assess Peripheral Edema.

Clin Med & Res ; 7(1-2):21-31.

Hedlund LW, Putman C.E, 1985. Methods for detecting

pulmonary edema, Toxicol Ind Health. 1(2):59-68.

Mayrovitz H.N, 2007. Assessing local tissue edema in

postmastectomy lymphedema, Lymphology 40: 87-94.

Pogue B.W. and Patterson M. S, 2006. Review of tissue

simulating phantom for optical spectroscopy, imaging

and dosimetry, JBO. 11(4), 041102.

Ohmae E., Yoshizawa N., Yoshimoto K., Hayashi M.,

Wada H., Mimura T., Suzuki H., Homma S., Suzuki N.,

Ogura H., Nasu H., Sakahara H., Yamashita Y., and

Ueda Y., 2018. Stable tissue-simulating phantoms with

various water and lipid contents for diffuse optical

spectroscopy, BOE 9, 5792-5808.

Kainz, W., Neufeld E., Bolch W., Graff C., Kim C. H.,

Kuster N., Lloyd B., Morrison T., Segars P., Yeom Y.,

Zankl M., Xu G., Tsui B.,. 2018, Advances in

Computational Human Phantoms and Their

Applications in Biomedical Engineering -A Topical

Review. IEEE Trans Rad and Plasma Med Scie,. 3(1):

1-23.

Saiko G, Zheng X, Betlen A, Douplik A, 2018.

Visualization of Methemoglobin Distribution in

Tissues: Phantom Validation, Adv Exp Med Biol 1072,

387-390

Braunwald E., 1994. Edema. In. Harrison’s Principles of

Internal Medicine, Isselbacher KJ, Braunwald E,

Wilson JD, et al., eds.

New York: McGraw-Hill; 13th

ed.

Bhave G, Neilson E.G , 2011, Body Fluid Dynamics: Back

to the Future, JASN, 22(12): 2166-2181.

Saiko G, Betlen A, 2019, Optimization of Band Selection

in Multispectral and Narrow-Band Imaging: an

Analytical Approach, Adv Exp Med Biol. (in press)

Dąbrowska A., Rotaru G.M., Spano F., Affolter Ch.,

Fortunato G., Lehmann S., Derler S., Spencer N.D.,

Rossi R.M., 2017. A water-responsive, gelatine-based

human skin model, Tribology Int, 113, 316-322.

Warner R., Myers M. C., and Taylor D. A., 1988. Electron

probe analysis of human skin: Determination of the

water concentration profile, J. Invest. Dermatol. 90,

218-224.

Saiko G., 2020. On the feasibility of skin water content

imaging adjuvant to tissue oximetry, Adv Exp Med Biol

(accepted).

BIOIMAGING 2020 - 7th International Conference on Bioimaging

134