In progress – view active session
Conference 13292
Photonics in Dermatology and Plastic Surgery 2025
25 - 26 January 2025
25 January 2025 • 8:40 AM - 10:30 AM PST
Session Chairs:
Hequn Wang, Johnson & Johnson Consumer Products (China), Mihaela Balu, Beckman Laser Institute and Medical Clinic (United States)
13292-31
longitudinal non-invasive optical biopsy of keratinocyte cancers to monitor efficacy and response to treatment
(Invited Paper)
25 January 2025 • 8:40 AM - 9:10 AM PST
Show Abstract +
Due to the impracticality of performing serial biopsies for the same lesion following treatments, efficacy data for low-risk keratinocyte skin cancer treatments is currently lacking. In this study, we employed non-invasive optical sectioning technology to monitor cellular dynamics and tissue architectural modifications post treatments every three months for up to a year. This optical sectioning was achieved utilizing a multimodal microscopy, integrating reflectance confocal microscopy (RCM), two-photon fluorescent (TPF) microscopy, and second harmonic generation (SHG) microscopy. Patients diagnosed with squamous cell carcinoma (SCC), basal cell carcinoma (BCC), and actinic keratosis (AK) were recruited in the study. Treatments evaluated included curettage and electrodessication (C&D), surgical excision, photodynamic therapy (PDT), cryotherapy, and topical chemo/immunotherapy.
13292-9
25 January 2025 • 9:10 AM - 9:30 AM PST
Show Abstract +
Certified multiphoton femtosecond laser tomographs have been used for skin imaging by (i) detection of autofluorescent coenzymes, melanin, and elastin, (ii) SHG of collagen, and (iii) two-photon FLIM since 2003. Besides detection of malignant melanoma, first MPT studies were used to measure long-term anti-ageing effects of skin care products by changing the elastin-collagen-ratio as well as short-term effects by influencing the NADH level with antioxidants. The penetration depth of ZnO sunscreen nanoparticles and microplastics in skin has been investigated. Nowadays, a multimodal ultracompact fiber laser tomograph (Prism Award Winner 2024) that includes confocal reflectance microscopy is employed by the cosmetic industry in Japan and France
Show Abstract +
Reflectance Confocal Microscopy (RCM) imaging shows cellular-level morphology of skin in vivo, noninvasively detects skin malignancies with high sensitivity (80-95%) and specificity (70-90%), reduces biopsies of benign skin lesions by 50% and is being routinely used at the bedside to guide patient care in clinical settings worldwide. However, clinicians using the RCM are unable to locate the field of view (FOV) in the context of the complete lesion being examined. We demonstrate feasibility incorporating wide-field (WF) imaging with RCM by integrating a micro-camera within the objective lens of a handheld confocal microscope. Preliminary clinical test results on human volunteers demonstrate the new handheld microscope’s ability to display a WF surrounding the location of RCM imaging on melanocytic and non-melanocytic skin lesions.
13292-22
25 January 2025 • 9:50 AM - 10:10 AM PST
Show Abstract +
We present an advanced multiphoton microscope for in-vivo dermal imaging of human skin with cellular-level resolution over a centimeter-square field of view. The large field of view facilitates easily locating permanent skin landmarks, such as pores, moles, skin folds, etc., ensuring consistent imaging at the same precise location over time. This capability enables systematic patient imaging over extended periods of time, potentially in the year scale, allowing time-lapse studies. We present examples of tracking cellular populations and collagen structures in the papillary and superficial dermis over a period of four months.
13292-41
25 January 2025 • 10:10 AM - 10:30 AM PST
Show Abstract +
Developing non-invasive imaging tools to enhance melanoma detection and diagnosis is crucial for improving patient survival and reducing healthcare costs. In this presentation, we discuss the results of a pilot study on ex-vivo and in-vivo MPM imaging of 20 patients with lesions suspicious for melanoma. We utilized our recently developed MPM clinical device, the Fast, Large Area Multiphoton Exoscope (FLAME), which captures in-vivo images over tissue areas up to 1 cm² with microscopic resolution in tens of seconds. FLAME employs a unique strip-mosaic (stage-mirror scanning) approach for in-vivo imaging. The label-free molecular contrast images are based on time-gated two-photon excited fluorescence of NADH/FAD, keratin, melanin, and elastin fibers and collagen's second harmonic generation.
Coffee Break 10:30 AM - 11:00 AM
25 January 2025 • 11:00 AM - 12:00 PM PST
Session Chairs:
Kristen M. Kelly, Beckman Laser Institute and Medical Clinic (United States), Kamran Avanaki, Univ. of Illinois Chicago (United States)
13292-44
25 January 2025 • 11:00 AM - 11:20 AM PST
Show Abstract +
The uses for a novel ablative Tm:YAP laser are presented. The laser can create ablative micro tunnels thinner than microns with control over the coagulation zone. These micro tunnels have been shown to have full reepithelization in 1 day, allowing short downtime for deep ablative treatment. These micro tunnels have been shown to increase the skin permeability of sumatriptan 42 times more than control samples, making the laser a good candidate for laser-assist drug delivery.
13292-20
25 January 2025 • 11:20 AM - 11:40 AM PST
Show Abstract +
Photo-sono therapy (PST) represents a novel anti-vascular approach based on cavitation-induced spallation. Currently, passive cavitation detection (PCD) is the prevalent technique for monitoring cavitation during treatment. However, PCD has the limitation of lacking spatial information about bubbles. To overcome this limitation, we developed an innovative cavitation mapping method by integrating Doppler optical coherence tomography (OCT) with PST. Doppler OCT is used to detect the extent of vascular removal by observing dynamic changes in cavitation during treatment, thus proving the treatment's efficacy. We first demonstrated the effectiveness of Doppler OCT in detecting cavitation within a vascular-mimicking phantom and compared these results to those obtained with a PCD and high-speed camera. Subsequently, the performance of the integrated system for treating microvessels was evaluated in vivo using rabbit ears. The findings indicate that real-time Doppler-OCT monitoring can improve the safety and efficiency of PST in removing microvasculature, thereby facilitating the development of individualized treatment strategies.
13292-5
25 January 2025 • 11:40 AM - 12:00 PM PST
Show Abstract +
Knowledge of the physical effects of pulsed dye laser (PDL) treatment of psoriatic lesions is essential in maximizing its disease modifying potential. Therefore, the main objective of this study was to provide estimates of these physical effects (for laser wavelengths of 585nm and 595nm), with the aim of identifying pathogenic processes which may be affected by these conditions. We modelled the laser light propagation and subsequent photothermal heating by numerically solving the transient diffusion and heat equations simultaneously. To this end, we used the finite element method in conjunction with an image-derived psoriatic lesion morphology (which was defined by segmenting blood vessels from a confocal microscopy image of a fluorescently-labelled section of a 3 mm punch biopsy of a psoriatic lesion). The resulting predictions of the generated temperature field within the lesion were then used to assess the possibility of stalling or arresting some suspected pathogenic processes.
Lunch Break 12:00 PM - 1:50 PM
25 January 2025 • 1:50 PM - 3:10 PM PST
Session Chairs:
Tashmeeta Ahad, The Univ. of British Columbia (Canada), Manu Jain, Memorial Sloan-Kettering Cancer Ctr. (United States)
13292-13
25 January 2025 • 1:50 PM - 2:10 PM PST
Show Abstract +
This research investigates the impact of repeated non-surgical aesthetic treatments, specifically Intense Pulsed Light (IPL) and Hyaluronic Acid (HA) injections, on skin morphology. Utilizing advanced photonics and bio-photonics methods like Optical Coherence Tomography (OCT) and VISIA imaging, the study analyzes both surface and internal skin changes. Results indicate that while both treatments induce short-term alterations in skin structure, these changes are transient and do not lead to significant long-term modifications. The study highlights the potential of these non-invasive imaging techniques for understanding subtle skin changes, particularly in evaluating the healing process and providing valuable insights for future diagnoses and efficacy assessments in medical aesthetics.
13292-23
25 January 2025 • 2:10 PM - 2:30 PM PST
Show Abstract +
Non-invasive imaging techniques, such as Skin Ultrasound, Optical Coherence Tomography (OCT), etc, offer researchers comprehensive understanding of skin qualities changes with certain depth profiles for skin beneath structure, which may reflect to skin surficial properties. This study investigates the skin quality changes in response to fractional CO2 laser therapy combined with cosmetic treatments. Using different imaging techniques, we aim to analyze and compare their effectiveness in correlating these changes to better understand the interactions between upper and deeper layers of skin.
13292-42
25 January 2025 • 2:30 PM - 2:50 PM PST
Show Abstract +
Traditional thermal modeling techniques, such as finite element analysis, for understanding the photothermal treatment of skin are complex, expensive, and time-consuming. We present a much simpler, but equally accurate approach using a lumped-element circuit model. This approach, in which discrete circuit elements represent thermal properties of the skin along with various heating and cooling elements, yields good physical insight into the spatial and temporal heat transfer. Free, off-the-shelf circuit analysis tools lend themselves remarkably well to this application, enabling rapid parametric analysis of different laser heating and cooling parameters.
13292-45
25 January 2025 • 2:50 PM - 3:10 PM PST
Show Abstract +
We describe a 1726 nm laser system for treating inflammatory acne. This system incorporates accurate, real-time temperature measurements of the skin being treated. The goal of any such laser treatment is to raise sebaceous gland temperature high enough to damage the gland while simultaneously sparing the surrounding dermal tissue from damage. To address this issue, we have developed a “treat to temperature” approach, which relies on a compact thermal imaging camera to provide real-time skin surface temperature measurements. Software, in turn, uses these measurements to control laser power and hit target pulse temperatures. We describe the engineering challenges with making consistently accurate temperature measurements.
Coffee Break 3:10 PM - 3:40 PM
25 January 2025 • 3:40 PM - 5:20 PM PST
Session Chairs:
Lise Lyngsnes Randeberg, Norwegian Univ. of Science and Technology (Norway), Hanna Jonasson, Linköping Univ. (Sweden)
13292-62
25 January 2025 • 3:40 PM - 4:00 PM PST
Show Abstract +
Spectral variations in human skin, especially across different skin tones, present significant challenges in accurately assessing skin characteristics such as oxygenation and moisture levels. Traditional assessment methods often produce inconsistent results due to the diversity in skin pigmentation, which affects the reliability of skin health diagnostics. Our study, which included 15 healthy individuals from three distinct skin color groups, investigated these spectral differences by analyzing reflectance in the visible and near-infrared (NIR) spectra, as well as skin autofluorescence under UVA and blue light excitation. To overcome these challenges, we developed the PIPA prototype, a handheld multimodal spectroscopy system that incorporates fluorescence spectroscopy. This device enhances non-invasive, multi-biomarker skin analysis, providing a more reliable and inclusive diagnostic tool suitable for all skin tones.
13292-61
25 January 2025 • 4:00 PM - 4:20 PM PST
Show Abstract +
Reduced epidermal pigment (melanopenia) is considered the hallmark pathophysiological difference between vitiligo and normal skin. The objective of this study is to quantitatively assess vitiligo in terms of its pathophysiological changes using multimodal spectroscopy non-invasively. Thirty-seven patients were recruited in this study, of which the vitiligo lesion and the adjacent normal skin were measured using diffuse reflectance and Raman spectroscopy. Skin color in CIE L*a*b* and biophysical properties including melanin, oxy-hemoglobin, deoxy-hemoglobin and scattering were quantified. It shows that vitiligo lesion has higher L*, lower a* and b* (p<0.0001); higher oxy-hemoglobin (p<0.0001) and oxygen saturation (p<0.0001), and lower melanin (p<0.0001), deoxy-hemoglobin (p=0.0014) and scattering (p<0.0001). Multimodal spectroscopy reveals that difference between vitiligo and its normal skin is more than melanin. There may be localized inflammation and photodamage as a consequence of reduced melanin protection within vitiliginous skin.
Show Abstract +
Skin bruises are the most common injury caused by physical violence, making them important indicators of ongoing abuse and evidence of past abuse. In court cases related to violence, expert witnesses are often asked to determine the age of bruise and to differentiate accidental and abusive bruises. This is a challenging task, since there is currently no objective, standardized method for documenting skin bruises. To address this, this project aims to develop an objective and accurate method for documenting skin bruises. Smartphone imaging, reflectance spectroscopy and hyperspectral imaging are used to document bruises of known origin.
13292-11
25 January 2025 • 4:40 PM - 5:00 PM PST
Show Abstract +
Acne is harder to diagnose darkly pigmented skin, leading to systemic underestimation of severity and undertreatment. However, acne lesions have increased sebum and tissue water content, both of which exhibit higher optical absorption in short-wave infrared wavelengths. SWIR imaging may assess acne equally well in lightly and darkly pigmented skin. We developed a SWIR multispectral system that acquires images at SWIR wavelengths and produces SWIR-pseudocolor images. We demonstrated that artificial sebum and water are distinguishable from each other in our system, and enrolled DPS subjects with acne and demonstrated a ten-fold contrast enhancement over visible photography. These results suggest that SWIR-MSI may transform practice by enhancing diagnostic accuracy and severity assessment of inflammatory acne in diverse populations.
13292-24
25 January 2025 • 5:00 PM - 5:20 PM PST
Show Abstract +
Melanoma, the most invasive skin cancer, has high mortality rates for advanced cases despite progress in treatment. For melanoma diagnosed at an early stage, the 5-year survival rate is above 90% highlighting the need for reliable, non-invasive diagnostics for both hyper- and hypopigmented melanoma. Cellular autofluorescence, influenced by NAD(P)H and flavins, is a valuable indicator of cell metabolism and a promising tool in this setting.
25 January 2025 • 7:00 PM - 9:00 PM PST
Every year at BiOS the community gathers at Saturday Night Hot Topics to hear the latest innovations in the biophotonics field. Don't miss this year's fast-paced program of world-class speakers. Open to all registered technical attendees.
26 January 2025 • 8:40 AM - 10:00 AM PST
Session Chairs:
Eric R. Tkaczyk, Vanderbilt Health One Hundred Oaks (United States), Aditi Sahu, Memorial Sloan-Kettering Cancer Ctr. (United States)
13292-15
26 January 2025 • 8:40 AM - 9:10 AM PST
Show Abstract +
The standard method for delineating skin cancer margins during surgery is histological examination of thin-sectioned and stained slides with several limitations. Optical tissue scanners have been developed to address the limitations of the current method by rapid examination without thin sectioning, but the imaging depth was low for 3D structure visualization. We recently developed open-top two-photon light sheet fluorescence microscopy (OT-TPLSFM) for high-throughput and high-depth imaging, achieving an imaging depth of 100 μm and a throughput of 0.24 mm²/s in human skin specimens. In this study, we applied OT-TPLSFM to various skin cancer specimens and conducted detailed analyses of cellular features for distinction among cancer cells, inflammatory cells, and normal glands. We trained and validated a classification model with mosaic 3D images, achieving 85% sensitivity and 95% specificity in identifying tumor regions. These results demonstrated the feasibility and efficiency of OT-TPLSFM for precise cancer detection during surgical procedures.
13292-39
intraoperative margin evaluation of fresh Mohs excisions using two-photon fluorescence microscopy
(Invited Paper)
26 January 2025 • 9:10 AM - 9:40 AM PST
Show Abstract +
We present interim results of a clinical trial evaluating an alternative workflow to intraoperative margin evaluation of fresh basal cell carcinoma Mohs specimens using two-photon fluorescence microscopy (TPFM). Fresh excisions are evaluated with TPFM and are then processed and evaluated with frozen section (FS) analysis. Afterwards, the frozen blocks are thawed and imaged with TPFM. We examine advantages of the TPFM workflow compared with FS with regards to timing, margin coverage, and absence of block facing leading to overtreatment.
13292-33
26 January 2025 • 9:40 AM - 10:00 AM PST
Show Abstract +
Mohs micrographic surgery (MMS) is the most effective therapy for treating non-melanoma skin cancers. Although effective, MMS is very labor intensive and time consuming, and requires access to highly trained dermatopathologists and histotechnicians, and on-site pathology lab. A technology that enables histopathological evaluation directly on the resected tissue will thus facilitate a broader implementation of MMS. Towards developing such technology, we report a novel inverted open-top fluorescence lifetime imaging (FLIM) system. The system enables simultaneous dual-excitation and multispectral FLIM imaging, targeting specific endogenous fluorophores relevant for discriminating cancerous cells from other nonmalignant skin tissue components. The imaging speed enables scanning a typical MMS resected tissue of 13x13 mm2 at a resolution of ~20 um within 4 minutes. We are currently building a library of FLIM and matching frozen section histology images of MMS resection tissue samples to identify FLIM biomarkers of skin cancer cells.
Coffee Break 10:00 AM - 10:30 AM
26 January 2025 • 10:30 AM - 11:50 AM PST
Session Chairs:
Ruikang K. Wang, Univ. of Washington (United States), Yusuke Hara, Shiseido Co., Ltd. (Japan)
13292-25
26 January 2025 • 10:30 AM - 10:50 AM PST
Show Abstract +
We demonstrate ultra-high-sensitive (UHS-) OCTA of in vivo human skin using a large number of frames per position (32 frames/position) and a very long (6.35-s) acquisition time window. The sample motion was restrained by a sample-fixation probe-tip and a motion correction algorithm.
The outer and inner forearms of 10 subjects were measured. Two dynamic OCT algorithms and complex-correlation OCTA algorithm were adopted for UHS-OCTA. Standard 4-frame OCTA was also measured.
Motion was well corrected for 7 subjects. UHS-OCTA provided very high contrast visualization of fine vessels. The vessel density of UHS-OCTA was significantly higher than standard OCTA (p < 0.0001).
13292-26
26 January 2025 • 10:50 AM - 11:10 AM PST
Show Abstract +
The development of treatments to reduce photodamage could be an important tool to reducing the prevalence of skin cancer. In this study, we investigate the use of optical coherence tomography (OCT) for assessing photodamage and the efficacy of photodamage prevention treatments. OCT images were acquired monthly on the back, side, and stomach of mice exposed to UVR over a period of 8 months. Two different systemic photodamage prevention treatments were tested: nicotinamide mononucleotide and polypodium leucotomos. Significant differences were observed in the total skin thickness (p=0.027) and optical attenuation coefficient (p<0.001) of the skin between all UVR-exposed mice and the non-exposed control group. Trends over time indicate that the skin-thickening and reduced attenuation occurred at a slower rate in both treatment groups compared with the UVR-exposed control. These results indicate that OCT could be a promising tool for monitoring photodamage in skin.
13292-36
26 January 2025 • 11:10 AM - 11:30 AM PST
Show Abstract +
Monitoring hair follicles is crucial for diagnosing and treating conditions like hair loss, skin diseases, and aging. Traditional histopathological methods, while standard, have limitations in providing dynamic and quantifiable data. This study presents a protocol utilizing a label-free, multi-scale optical imaging device combining Optical Coherence Tomography (OCT) and Optical Coherence Microscopy (OCM). This approach allows real-time, 3D morphological monitoring and rapid quantification of hair follicles, offering histopathological insights within an hour. The integration of in vivo/ex vivo imaging provides comprehensive functional and structural information, enhancing applications in histopathology, particularly in surgical settings, and reducing costs through sample reuse.
13292-40
26 January 2025 • 11:30 AM - 11:50 AM PST
Show Abstract +
Adipose tissue (AT) undergoes dynamic remodeling in response to differential nutritional supply. Changes in AT morphology accompany its functional changes, and imaging is a powerful tool for studying and understanding AT remodeling. Histopathological analysis conventionally used for studying AT remodeling requires tissue removal and processing, providing only a single time-point measurement. We developed a workflow capable of carrying out repeated measurements for longitudinal monitoring of the expansion and reduction of intact AT. Non-invasive real-time imaging of live mice was carried out to obtain stable scans which are then analyzed using our semi-automated labeling algorithm to obtain quantitative readouts. We investigated adipose tissue remodeling in mice of various metabolic conditions, such as obesity and hypercholesterolemia. Through this, we demonstrate AT live imaging as a promising tool for reflecting changes in metabolic status.
Lunch Break 11:50 AM - 1:20 PM
26 January 2025 • 1:20 PM - 2:40 PM PST
Session Chairs:
Anthony J. Durkin, Beckman Laser Institute and Medical Clinic (United States), Rolf B. Saager, Linköping Univ. (Sweden)
13292-21
26 January 2025 • 1:20 PM - 1:40 PM PST
Show Abstract +
Assessing burn severity is crucial for reducing scarring and infection risk, but current visual and tactile methods can be subjective. A more objective method using spatial frequency-domain imaging (SFDI) and machine learning (ML) has shown promise. This technique was tested on a swine model, collecting data with two SFDI systems, the Ox-Imager RS and Clarifi (Modulim Inc.), which use different wavelengths and spatial frequencies. Previously, this data was used to train support vector machine (SVM) models. This study evaluates ensemble learning (EL) algorithms, specifically random forest and Libra, against SVM for classifying burn severity in pigs. The methodology ensures fair comparison and avoids overfitting. Results show EL models can improve performance by up to 50% over SVM in metrics like F1-Score, producing more accurate classification maps with fewer false positives, and identifying areas likely to heal, offering practical clinical insights.
13292-37
26 January 2025 • 1:40 PM - 2:00 PM PST
Show Abstract +
Spatial frequency domain imaging (SFDI) is an optical technique for measuring optical properties quantitatively using structured illumination. In clinical settings, SFDI finds numerous applications, from cancer detection to monitoring pressure ulcers. Because of the non-contact, non-invasive nature of the technique, it is especially suitable for the assessment of wounds.
When performing measurements on wounds, a 2-layer model is preferred over homogeneous models because of the presence of a thin superficial layer. In order to detect optical contrast in thin layers, a technique that can perform depth-resolved measurements is necessary.
Multi-depth SFDI is a new approach using the difference in the penetration depth of modulated light, depending on its spatial frequency. The technique interprets the depth-resolved data according to a 2-layer model, to quantify scattering contrast in thin layers of tissue.
In this work, we validate the multi-depth SFDI approach using thin silicone tissue-mimicking phantoms of different thickness.
13292-47
26 January 2025 • 2:00 PM - 2:20 PM PST
Show Abstract +
Burn wound treatment varies with severity. Superficial partial-thickness (SPT) burns typically don't need grafting, while deep partial-thickness (DPT) and full-thickness burns often do to prevent infection, scarring, and contraction. Early intervention is crucial, but physician accuracy in burn assessment is 60-80%. Thus, a reliable quantitative measure is essential for optimal patient outcomes. A portable imager is being developed to meet this need using Spatial Frequency Domain Imaging (SFDI), a non-invasive multi-spectral imaging technique. This study compares machine learning models for burn classification. Burns on a porcine model were imaged with SFDI and photography, and biopsies were collected, from pre-burn to 28 days post-burn. Two classifiers were developed, with the region-based model outperforming the histology-derived model. Based on these results, researchers should consider burn heterogeneity in model development.
13292-38
26 January 2025 • 2:20 PM - 2:40 PM PST
Show Abstract +
Scleroderma, or systemic sclerosis (SSc) is a chronic, autoimmune disorder that manifests in the fibrosis of skin and internal organs. The clinical gold standard used to track SSc disease progression is the modified Rodnan skin score (mRSS), based on clinical palpation. We have been investigating Spatial Frequency Domain Imaging (SFDI), a widefield, non-contact diffuse optical imaging technology as an alternate quantitative and objective method to track progression of SSc by measuring tissue reflectance and optical properties. To date we have performed SFDI measurements on 33 SSc patients and 29 healthy controls, with 10 patients and 3 controls being measured longitudinally. We found that there were proportional changes in SFDI metrics (μs' (851 nm) and Rd (851 nm, 0.2 /mm)) corresponding to SSc progression assessed by mRSS, ultrasound-measured dermal thickness, durometry scores and histopathological metrics, suggesting SFDI could provide an improved method to track SSc progression.
Coffee Break 2:40 PM - 3:10 PM
26 January 2025 • 3:10 PM - 4:30 PM PST
Session Chairs:
Conor L. Evans, Wellman Ctr. for Photomedicine (United States), Kivanc Kose, Memorial Sloan-Kettering Cancer Ctr. (United States)
13292-19
26 January 2025 • 3:10 PM - 3:30 PM PST
Show Abstract +
This presentation explores the latest advancements in photonics technologies applied to dermatology and plastic surgery. We delve into cutting-edge imaging techniques, enhanced histological analysis, and the integration of artificial intelligence to improve diagnostic accuracy and treatment outcomes. Additionally, we highlight innovations in minimally invasive in vivo sampling and imaging, which offer unprecedented insights into skin and tissue health with reduced patient discomfort. Join us to discover how these technological breakthroughs are revolutionizing clinical practices and paving the way for more precise, efficient, and patient-friendly dermatological and surgical interventions.
13292-34
26 January 2025 • 3:30 PM - 3:50 PM PST
Show Abstract +
This study enhances lesion analysis and understanding of skin cancer progression by simulating potential malignant transformations of benign nevi. Using CycleGAN and optical flow analysis, we generated detailed sequences of lesion changes and evaluated them against ABCDE rule metrics. The simulations offer valuable insights for early detection and better understanding of melanoma growth, providing significant advancements in dermatological diagnostics and patient education.
Show Abstract +
Reflectance confocal microscopy (RCM) improved the diagnosis of basal cell carcinoma (BCC) but it’s actual utility over dermoscopy for equivocal lesions has not been evaluated. Artificial intelligence (AI) algorithms have been developed to aid automated detection of BCC on RCM, however, as diagnosis in clinics relies on dermoscopy and RCM, a combined algorithm would be valuable for equivocal lesions.
We performed a multi-reader study to assess the utility of RCM over dermoscopy for diagnosis of equivocal lesions for BCC. We developed a combined dermoscopy-RCM AI algorithm and compared it the human reader performance.
251 equivocal BCC lesions were evaluated by 15 readers. They first read dermoscopy images and gave a diagnosis of BCC or not-BCC along with confidence level, followed by RCM images. A final diagnosis was made on dermoscopy and RCM images. A change in confidence was also noted over dermoscopy. Two separate Convolutional Neural Networks using dermoscopy and RCM images were developed. They were combined and tested on the reader dataset. A combined AI algorithm (RCM + dermoscopy) can further assist in the automated diagnosis of BCC, beneficial for novices.
13292-113
26 January 2025 • 4:10 PM - 4:30 PM PST
Show Abstract +
Automated dermatological diagnosis with artificial intelligence has made significant progress in recent years, especially with deep learning algorithms. However, classifying rare and unseen skin lesion types remain a challenge mainly due to the limited availability of annotated data. This paper proposes a new approach that combines transfer learning algorithm with zero-shot learning using the HAM10000 dataset. A pre-trained Convolutional Neural Network model extracts features from dermatoscopic images, fine tune it on a subset of common lesion types in the dataset and then use zero-shot learning principles to infer diagnosis for rare or unseen lesion types. By using semantic embeddings and similarity metrics, the model compares embeddings of unseen lesions with prototypes of known lesion types. The model achieves an overall accuracy above 85% on the test set and demonstrates the potential of combining transfer learning with zero-shot learning to improve the accuracy, generalization and clinical utility of automated dermatological diagnosis systems.
26 January 2025 • 4:30 PM - 5:30 PM PST
Session Chairs:
Bernard Choi, Beckman Laser Institute and Medical Clinic (United States), Boris Majaron, Jožef Stefan Institute (Slovenia)
13292-48
26 January 2025 • 4:30 PM - 4:50 PM PST
Show Abstract +
This presentation will describe a color polarization camera system designed to quantify blood volume and tissue oxygen saturation within superficial and deep layers of human skin. The system captures co- and cross-polarized color (RGB) images in a single acquisition frame. Polarization information is used to discriminate between skin layers and RGB color information is used as input to a neural network to predict tissue blood volume and oxygen saturation. This presentation will describe the technical implementation of the method and demonstrate its performance on in vivo skin from human subjects with a range of skin tones and ages.
13292-49
26 January 2025 • 4:50 PM - 5:10 PM PST
Show Abstract +
Objective skin color measurements are useful to measure disease progression and ensure clinical trials enroll diverse participants. Colorimeters derive color measurements in the L*a*b* color space but can be difficult to use at certain sites. The L*,b* plane is useful for calculating the Individual Typology Angle (ITA), which is used to quantify melanin content. Photography has been considered a tool for colorimetry but has been limited in maintaining color consistency. Therefore, we utilized a dermatoscope, a polarized 10x magnified epiluminescence imaging device, with a 9.5x5.6mm 30-color calibration target for color correction. For 23 subjects with various pigmentation, ITA from -52° to 65°, the ITA was calculated from corrected and uncorrected images and compared to ITA from a Konika Minolta CM700d spectrophotometer. The mean error for corrected ITA values was lower than uncorrected and had a stronger correlation to measured ITA indicating the feasibility of color-corrected dermatoscope as a colorimeter.
13292-59
26 January 2025 • 5:10 PM - 5:30 PM PST
Show Abstract +
We investigated a change in perfusion of pig burns of different severities over the course of 3 days with the application of standard burn care including field debridement. Speckle Flow Index (SFI) values for each burn duration are visibly different starting on day 0 and show significant differences on day 1 and 3. Additionally, linear regression was applied to create burn depth maps from SFI and graded biopsy data
26 January 2025 • 8:00 AM - 9:30 AM PST
13292-3
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
Skin aging is induced by ultraviolet (UV) radiation, hormonal deficiencies, and environmental factors, resulting in a decrease in skin thickness, collagen, and elastin levels. Among various medical and surgical treatments, laser therapy has become increasingly popular as a non-invasive treatment for skin rejuvenation. The aim of this study is to demonstrate the effect of laser treatment with triple wavelength on skin rejuvenation in an in vivo model in terms of histological and molecular analysis. The degree of skin rejuvenation was measured at 0, 7, 14, and 28 days after laser treatment. The current study demonstrated that the proposed laser system could be effective for skin rejuvenation by delivering three wavelengths (755, 808, 1064 nm) simultaneously to obtain a synergistic effect.
13292-35
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
This study introduces a high-resolution imaging module for intelligent total body scanners aimed at early melanoma detection. By utilizing a focus stacking approach and deep learning-based super-resolution techniques, the system produces high-quality, non-contact dermoscopic images. This method ensures comprehensive focus across skin topographies, significantly enhancing diagnostic accuracy. The approach marks a substantial improvement in non-invasive skin imaging, providing an all-in-focus view essential for accurate melanoma detection.
13292-46
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
Lichen sclerosus (LS) is a chronic inflammatory condition primarily affecting the anogenital skin in women, with unclear prevalence due to underdiagnosis. Estimated to affect up to 3% of postmenopausal women, LS often faces diagnostic delays, leading to severe health outcomes like scarring, infections, and pain. Skin biopsies, the standard diagnostic method, are invasive and time-consuming. About 5% of LS patients develop vulvar squamous cell carcinoma (SCC). Given these challenges, we have developed a 1.7-μm polarization-sensitive optical coherence tomography technique for noninvasive, real-time diagnosis and monitoring of LS lesions, enabling early detection of malignant transformations and better therapeutic monitoring.
13292-50
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
The dermal epidermal junction (DEJ) is an important indicator of skin health. Optical coherence tomography (OCT) allows noninvasive visualization of skin and the DEJ, but the training required for users serves as an obstacle. Delineation and disease detection algorithms can mitigate this issue. After creating DEJ segmentation machine learning models, they were modified for basal cell carcinoma (BCC) detection, the most common non-melanoma skin cancer. Two approaches of detection were taken: 1) Use original machine learning models to segment DEJ and classify skin based on DEJ linear length density and 2) create unsupervised models specifically for classifying image columns as containing BCC or not. With 27 healthy subjects (1,215 images) and 6 BCC subjects (446 images), both approaches yielded receiver operating characteristic (ROC) curves with areas greater than 0.9, indicating that both are excellent approaches for BCC classification.
13292-52
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
This study introduces a novel photoacoustic ultrasound probe for high-resolution skin imaging, focusing on accurate melanoma thickness measurement. The probe features a compact linear array transducer with 256 elements operating at 44 MHz. It employs a coaxial illumination system using thin chromium films within a gelatin block attached to the transducer by a 3D holder. This setup efficiently transmits ultrasound waves and reflects light, optimizing tissue illumination and reducing out-of-plane artifacts. This compact, handheld probe integrates ultrasound and photoacoustic imaging, offering significant advancements in melanoma visualization and thickness measurement for improved diagnosis and treatment planning.
13292-53
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
Laser illumination is crucial in photoacoustic imaging and must meet safety regulations to avoid skin hyperthermia. Skin characteristics significantly impact heat absorption during laser exposure. This study assesses the thermal effect of laser light on skin, focusing on parameters such as skin type, age, and gender. A robust bioheat model predicts skin temperature changes due to laser-tissue interactions. Results show that 100 Hz frequency causes a higher temperature increase than 50 Hz within 6 seconds of exposure. Older individuals exhibit a less pronounced temperature increase than younger individuals, highlighting the need for age-specific laser treatment protocols to ensure safety and effectiveness.
13292-54
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
This research explores using optical coherence tomography (OCT) to diagnose nail unit melanoma (NUM) non-invasively, focusing on potential differences in OCT images across skin tones (Fitzpatrick I-IV). OCT offers a less invasive, more accurate alternative to biopsies, which are difficult due to nail anatomy and can result in delayed diagnosis and poor prognosis. This method could improve NUM diagnosis for a variety of skin tones.
13292-55
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
Speckle noise is a common issue in optical coherence tomography (OCT) imaging systems, resulting from random interference among coherent reflections from scatterers. This granular noise degrades image quality, posing challenges for image restoration. This paper presents practical noise model for OCT images by analyzing speckle distribution derived from statistical features in the difference between single OCT images and co-registered averaged images. The model combines Rayleigh, Gamma, Nakagami, and K-Distributions, with weighting parameters determined via Maximum Likelihood Estimation. This model is applied to reduce speckle noise in OCT skin images, evaluated both qualitatively and quantitatively across different skin types and ages.
13292-56
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
We introduce a novel approach to integrate Fitzpatrick Skin Type (FST) classification with Optical Coherence Tomography (OCT) to enhance dermatological assessment. By imaging three individuals from each of the six Fitzpatrick skin types across 20 body regions, we quantified optical coefficients—absorption and scattering properties—from OCT images. These coefficients, indicative of skin's structural characteristics, are analyzed to identify correlations with specific skin types. Integrating OCT with FST classification minimizes subjective assessments and provides more precise skin type identification, improving dermatological care and ensuring fair healthcare outcomes for diverse groups.
13292-63
26 January 2025 • 8:00 AM - 9:30 AM PST
Show Abstract +
Early detection of pressure injuries is crucial due to their negative impact on both patients and the healthcare system. This study employs spectroscopy and Partial Least Squares Regression (PLSR) to estimate skin moisture content as a biomarker for the early detection of pressure injuries. Moisture was measured at six points on each hand (one moisturized and one dry) using the MoistureMeterEpid, SEM Scanner, and two spectrometers (PIPA, VIAVI), with participants of varying skin tones. The data underwent preprocessing with Standard Normal Variate (SNV), and the PLSR model was trained to predict moisture levels from the spectrometer data using the reference measurements.
This conference is not accepting post-deadline submissions.