This PDF file contains the front matter associated with SPIE Proceedings Volume 12816, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

We demonstrated dynamic optical coherence tomography (D-OCT) to human skin in vivo by applying a sample fixation attachment and bulk motion correction algorithm to correct the motion artifacts. A D-OCT contrast of logarithmic intensity variance (LIV) was calculated. Without sample fixation attachment and the motion correction algorithm, the whole image area exhibited high LIV, and no meaningful structure was seen. The application of the motion correction methods revealed fine en face vessel structures, which cannot be seen in OCTA. The statistically significant motion artifact reduction capability of our motion correction method was also shown by t-test.

Port-Wine Birthmarks (PWB) are vascular birthmarks occurring in 0.3-0.5% of the population.1 First-line treatment is pulsed-dye laser (PDL), however, there is rarely full clearance of lesions. Dynamic-OCT (D-OCT) was used to measure blood vessel density (%) and diameter (micrometers) in the dermis of single PWBs at depths between 0.15mm and 0.5mm. Average ratios of vessel density and diameter in affected compared to control skin were obtained for each PWB by averaging data for all spots within a lesion. Density was consistently greater in affected skin compared to control skin at deeper depths in hypertrophic lesions than in flat lesions. Diameter did not vary consistently with depth. In conclusion, there is a great deal of variability in OCT-measured vessel density and diameter within and between PWBs. OCT provides data on individual lesion characteristics that could potentially be used for laser treatment planning.

Effective burn wound management is informed by accurate severity assessment. Superficial partial-thickness burns do not require surgical intervention, while deep partial-thickness and full-thickness burn wounds necessitate skin grafting to minimize infection, contraction, and hypertrophic scarring. Visual-tactile wound assessment is subjective and error-prone, especially for inexperienced practitioners. A field- and hospital-deployable device, capable of quantifying both extent and severity of burns, could enable rapid, objective burn severity measurement with commensurate improvement in patient outcomes. Our group has previously shown that spatial frequency domain imaging (SFDI), a non-invasive, wide-field optical imaging technique, can accurately assess burn wound in a porcine model of controlled, graded burn severity^{[1, 2]}. The device employed (OxImager RS) eight modulated wavelengths and five spatial frequencies and the classification of severity relies heavily on reduced scattering coefficient (tissue microstructure)^[3]. In the work that we present here, we demonstrate the burn severity prediction performance of a dramatically streamlined version of SFDI that employs a single modulated wavelength in addition to five unmodulated wavelengths. This device, known as Clarifi (Modulim, Irvine CA), is currently in refinement for ruggedization and usability for a variety of situations in which the environment is more demanding than hospital clinics. In addition, we have developed a machine learning model capable of categorizing burn severity in a porcine model of graded burns using a reduced dataset of unprocessed calibrated reflectance images generated by the device. Outputs of the model are designed to be easily interpretable and clinically actionable, exhibiting a pixelwise cross-validation accuracy of up to 99%.

The concept of photothermal tomography (PTT) involves spatially and temporally resolved detection of blackbody emission from the sample surface after irradiation with a short light pulse. In principle, this allows reconstruction of the light-induced temperature field inside the sample, thus enabling three-dimensional imaging of absorbing structures in strongly scattering biological tissues and organs. However, development of accurate and robust PTT methodology has proven difficult due to the large size and severe ill-posedness of the underlying inverse problem, aggravated by the inherently low signal-to-noise ratios of active infrared radiometry. We discuss here our recently developed PTT system and its first application to human skin in vivo. The experimental setup involves a medical-grade laser emitting milisecond pulses at 532 nm, and a fast mid-infrared camera equipped with a microscope objective. A custom code written in Python is used to reconstruct three-dimensional images of the absorbing structures by performing iterative multidimensional minimization of the difference between the analytically predicted and experimental radiometric record, using a projected v-method algorithm. The described approach produces a rather sharp and high-contrast tomographic image of a tattoo layer in a human volunteer’s skin with the onset depth of ~0.15 mm. No evident artifacts or even noise appear elsewhere in the reconstructed volume. Applying quadratic binning of the radiometric record significantly reduced the computational load, enabling us to reconstruct a volume of 5.6 x 2.8 x 0.9 mm³ with nominal resolution of 25 x 25 x 15 μm³ using a personal computer in less than one minute.

Multiphoton tomography is a clinical imaging method to obtain high-resolution optical biopsies of human skin and to perform optical metabolic imaging (OMI) by time-correlated single photon counting / fluorescence lifetime imaging (FLIM) of autofluorescent coenzymes. We report on a long-term MPT-OMI study on two volunteers during oxygen treatment. Metabolic changes of epidermal skin cells have been recorded during daily two hours oxygen inhalation over an one-week-treatment period. Multimodal MPT (confocal reflection, autofluorescence, FLIM, SHG) has been performed with the tomograph MPTcompact based on a compact femtosecond fiber laser located inside a 360° imaging head mounted on a flexible mechanical arm. Imaging/tracking of specific intratissue cells could be performed with submicron resolution over two hours during oxygen inhalation. Oxygen inhalation influences the ratio free to bound NADH mainly in the basal cell layer close to the capillaries compared with the stratum granulosum. In conclusion, multiphoton tomography can be used for in vivo metabolic time-lapse high-resolution imaging of human skin during treatment, such as oxygen therapy.

We report a quantitative tissue imagining method that combines spatial frequency domain imaging (SFDI) and near-infrared spectroscopy (NIRS) by using an illumination of multiple scanned lines. SFDI is suitable for measuring superficial and shallow tissues (<a few mm) with sinusoidal or stripe illuminations while NIRS, that measures the points apart from a point illumination, is for deeper tissues (a few to tens mm). Our scheme performs these methods by one measurement and provide depth-dependent reflectance images for shallow and deep tissues. At every pixel, a series of the pixel values for all scanning steps is processed in two different ways: SFDI by fast Fourier transform and NIRS by finding the minimum value. We assembled a prototype system and measured a phantom and a human arm. The line pitch and wavelength were 10mm and 785nm, respectively. We obtained reflectance images from the 32 captured images using the SFDI processing for 0-0.4mm-1 and the NIRS processing for the line pitches of 0.625- 10mm. The images by SFDI for higher spatial frequencies depicted the shallower blood vessels. In those by NIRS, the contrast of the deeper blood vessels was enhanced for the larger line pitches.

We previously developed a pre-vascularized three-dimensional (3D) cultured human skin model, for which cells were cultured on a porous polymer membrane to supply a culture medium to the tissue. The model was transplanted onto a full-thickness skin defect in a mouse, showing the efficacy as a skin graft. However, there were difficulties in separating the cultured skins from the membranes, as well as in handling the separated soft skins for transplantation. To solve these problems, we recently developed biodegradable porous membranes that enable skin grafting together with the membranes. Poly (lactic-co-glycolic acid) (PLGA) thin films were irradiated with femtosecond laser pulses to create micro through-holes to produce porous membranes. The membrane showed complete decomposition in the mouse subcutaneous tissue within 35 days after implantation. Three-dimensional skins cultured on the membranes were then transplanted together with the membranes onto skin defects in mice, showing reepithelization in the grafted tissues. However, decomposition of the membrane was limited at the early-stage post-transplantation, and insufficient engraftment was observed in some cases. In this study, we increased the hole density of the PLGA membranes to improve their decomposition rate in tissue. We observed that the membrane was completely decomposed in the mouse subcutaneous tissue within 24 days after implantation. On the membranes, 3D skins containing vascular networks were cultivated. However, we encountered another problem that the mechanical strength of the membranes decreased with the increased hole density; some membranes were torn during skin cultivation. Considering this, we plan to further optimize the membrane conditions next.

Picosecond (ps) lasers have emerged as a promising tool for effectively removing unwanted tattoos. However, data on the in vivo elimination process after ps laser treatment is limited. Multiphoton tomography (MPT) is an established intravital imaging technique and provides high-resolution subcellular visualization of the morphology of tattoos and the skin. The aim of this study is to visualize and characterize the distribution of tattoo particles after ps laser treatment using MPT-FLIM. Patients seeking a laser tattoo removal were recruited and received one to three 532/1064nm ps laser treatment sessions. MPT-FLIM measurements were scheduled six to eight weeks post-treatment. The number and size of tattoo granules decreased with subsequent treatments. Even months after the last ps laser treatment, tattoo pigments were still detectable in the upper epidermal layers and around vessels, indicating a continuous elimination process. Metabolic analysis by calculating the mean fluorescence lifetime (τ_m) revealed a persistent inflammatory status at follow-up visits. Our study concludes that MPT-FLIM is a suitable non-invasive technique for characterizing tattoo particles after ps laser treatment. Its clinical application enables monitoring and adjustment of treatment based on individual responses.

The Reflectance Confocal Microscopy – Optical Coherence Tomography (RCM-OCT) device has demonstrated its effectiveness in the in vivo detection and depth assessment of basal cell carcinoma (BCC), though its interpretation can be challenging for novices. Artificial intelligence (AI) has the potential to assist in identifying BCC and measuring its depth in these images. Our goal was to develop an AI model capable of generating 3D volumetric representations of BCC to enhance its detection and depth measurement. We developed AI models trained on OCT images of biopsy-confirmed BCC to detect BCC, generate 3D volumetric representations, and automatically assess tumor depth. These models were then tested on a separate dataset containing images of BCC, benign lesions, and normal skin. The effectiveness of the AI models was evaluated through a blinded reader study and by comparing tumor depth measurements with those obtained from histopathology. The addition of AI-generated 3D renders of BCC improved BCC detection rates, with sensitivity increasing from 73.3% to 86.7% and specificity from 45.5% to 48.5%. A Pearson Correlation coefficient r² = 0.59 (p=0.02) was achieved in comparing tumor depth measurements between AI -generated renders and histopathology slides. Incorporating AI-generated 3D renders has the potential to improve the diagnosis of BCC and the automated measurement of tumor depth in OCT images, reducing reader dependent variability and standardizing diagnostic accuracy.

Deep learning, AI and machine learning are emerging as important tools e.g., to segment, classify and detect pathologies medical diagnostics. Powerful and easy to use frameworks for machine learning have increased the accessibility of these methods. At the same time this removes the need for machine learning experts with deep insights in the limitations of the technology and places the power of AI in the hands of domain experts. Although such high-risk domains are very diverse, the challenges using machine learning are related. This paper will discuss insights from application of machine learning in domains as diverse as medical image analysis, and conditionbased monitoring in the Norwegian Oil and gas industry. We then discuss how these insights can be applied when analyzing hyperspectral data from human skin.

Polarization-Sensitive Optical Coherence Tomography (PS-OCT) is a real-time 3D imaging technique providing structural and functional contrasts in tissue. Previously, we have quantified multiple contrasts such as back-scattered intensity, accumulated phase retardation, local birefringence, and degree of polarization uniformity in the epidermis and dermis of the skin. Here, we add the Attenuation Coefficient (AC) contrast to quantify light attenuation from absorption and scattering in skin layers. Two techniques are utilized to obtain AC. One involves slope fitting to the logarithm of intensity A-lines, and the other uses a depth-resolved model-based reconstruction of AC. To investigate the effect of skin tone on AC, we first used skin phantoms with various absorption and scattering coefficients. It is found that the color of the phantom directly correlates with the absorption coefficient, while AC correlates with the sum of the absorption and reduced scattering coefficient. Darker objects, indicative of a higher concentration of melanin in the skin tissue, show a higher absorption coefficient. AC is further analyzed on in vivo skin imaging with different skin tones. However, no clear correlation between AC and skin tone is observed. This lack of correlation is likely due to the absorption coefficient being much smaller than the scattering coefficient in skin tissue at the OCT imaging wavelength of 1060nm.

Our group has recently presented a novel technique for noninvasive assessment of the structure and composition of human skin in vivo. The approach combines two optical techniques, photothermal radiometry and diffuse reflectance spectroscopy in visible part of the spectrum with numerical modeling of light transport in a four-level model of human skin, and enables assessment of the contents of specific chromophores (e.g., melanin, oxy-, and deoxy-hemoglobin), as well as scattering properties and thicknesses of the epidermis and dermis. In this study, we try to quantitatively validate the described approach by analyzing a series of optically homogeneous skin phantoms with varying concentrations of absorbing dye (Congo red) and optical scatterers. Twelve such phantoms were made from bovine collagen gel and their scattering properties were controlled by adding TiO₂ nanoparticles with diameters of 200nm and 490nm, aiming at the range of values typical of human dermis. The applied Monte Carlo model of light transport takes into account the actual diameter and thickness of the phantoms and allows simultaneous assessment of the absorption coefficient as well as the amplitude and spectral power of the reduced scattering coefficient. The obtained values are proportional to the respective concentrations of the absorbing dye and scatterers, but don’t match the theoretical predictions (from a Mie simulator).

Histologic examination of skin biopsies is currently the gold standard to definitively diagnose malignant skin lesions; however, biopsies are minor, invasive procedures with potential risks. With the advancement of imaging techniques such as laser speckle contrast imaging (LSCI), it is now possible to evaluate neoplastic skin lesions in real-time and noninvasively. LSCI has been widely used to image surface blood flow in tissues, such as skin, retina, and brain. In this preliminary study, we hypothesized that blood flow within microvessels differs between neoplastic and non-neoplastic skin. This study presents a descriptive demonstration of LSCI application in dermatology. LSCI was utilized to assess surface blood flow in potentially neoplastic skin lesions at our institution’s dermatology clinics. Preliminary data demonstrated decreased contrast within speckle contrast images of malignant and premalignant skin lesions, suggesting increased blood flow to these areas of interest. LSCI may show utility as a noninvasive technique to evaluate neoplastic skin lesions prior to biopsy; however, further systematic optimization is required.

In recent years, ultraviolet (UV) light therapy has been attracting attention as a treatment method for skin diseases such as atopic dermatitis, vitiligo, and psoriasis which have been considered difficult to treat. However, regarding the research on treatment, clinical studies have been the main research focus without much in-depth consideration into the radiative properties of UV light within the skin. Therefore, it is essential to know the optical properties, scattering and absorption coefficients, of the skin in the UV wavelength range to improve the treatment strategy. Our group has been developing a non-invasive method, reflection spatial profile method (RSPM), for measuring optical properties of skin and elucidated optical property differences arising out of cancer drugs[1]. In this study, we have developed a novel system that operates in the range of UV wavelengths irradiating the skin with a structured incoherent source and detecting the reflected light with a CCD camera. Measurements are being conducted with human subjects of different age groups. Based on results obtained from the forearm and hand measurements of 76 subjects consisting of 31 males and 45 females in their 20s~30s. UV light penetration depth can be estimated from the optical properties obtained. It was found that the amount of light absorption can vary up to 1.5 times at the same depth depending on gender, age, and site.

The position and appearance of the dermal epidermal junction (DEJ) is an important indicator of skin health. Optical coherence tomography (OCT) is used for noninvasive skin imaging but is impeded by the training required. DEJ delineation algorithms address this issue but with limited consideration of diversity in samples. In this study, marked images of a variety of body regions, age groups, and Fitzpatrick skin type (FST) groups were used. To find the DEJ automatically, image columns were matched to a body region-specific swatch of similar columns based on cosine similarity. Our results demonstrated the swatches method could determine around 87% of all automatic markings within 39 micrometers axially of the manually identified DEJ.

This study introduces a learning-assisted denoising technique for skin Optical Coherence Tomography (OCT) images. By combining Reinforcement Learning (RL) with the Denoising Convolutional Neural Network (DnCNN), we achieve enhanced denoising capabilities. The method iteratively refines DnCNN parameters through RL-guided policies, demonstrating superior performance. Tailored for skin OCT images, the approach prioritizes preserving vital structures for accurate clinical assessments. This integration of RL into DnCNN training represents a promising advancement in medical image denoising, particularly for dermatological diagnostics.