Automated Segmentation of Skin Strata in Reflectance Confocal Microscopy Depth Stacks

Evaluating facial dermis aging in healthy Caucasian females with LC-OCT and deep learning

Recent advancements in high-resolution imaging have significantly improved our understanding of microstructural changes in the skin and their relationship to the aging process. Line Field Confocal Optical Coherence Tomography (LC-OCT) provides detailed 3D insights into various skin layers, including the papillary dermis and its fibrous network. In this study, a deep learning model utilizing a 3D ResNet-18 network was trained to predict chronological age from LC-OCT images of 100 healthy Caucasian female volunteers, aged 20 to 70 years. The AI-based protocol focused on regions of interest delineated between the segmented dermal-epidermal junction and the superficial dermis, exploiting complex patterns within the collagen network for age prediction. The model achieved a mean absolute error of 4.2 years and exhibited a Pearson correlation coefficient of 0.937 with actual ages. Furthermore, there was a notable correlation (r = 0.87) between quantified clinical scoring, encompassing parameters such as firmness, elasticity, density, and wrinkle appearance, and the ages predicted by deep learning model. This strong correlation underscores how integrating emerging imaging technologies with deep learning can accelerate aging research and deepen our understanding of how alterations in skin microstructure are related to visible signs of aging.

Automating reflectance confocal microscopy image analysis for dermatological research: a review

SIGNIFICANCE

Reflectance confocal microscopy (RCM) is a noninvasive, in vivo technology that offers near histopathological resolution at the cellular level. It is useful in the study of phenomena for which obtaining a biopsy is impractical or would cause unnecessary tissue damage and trauma to the patient.

AIM

This review covers the use of RCM in the study of skin and the use of machine learning to automate information extraction. It has two goals: (1) an overview of information provided by RCM on skin structure and how it changes over time in response to stimuli and in disease and (2) an overview of machine learning approaches developed to automate the extraction of key morphological features from RCM images.

APPROACH

A PubMed search was conducted with additional literature obtained from references lists.

RESULTS

The application of RCM as an in vivo tool in dermatological research and the biologically relevant information derived from it are presented. Algorithms for image classification to epidermal layers, delineation of the dermal-epidermal junction, classification of skin lesions, and demarcation of individual cells within an image, all important factors in the makeup of the skin barrier, were reviewed. Application of image analysis methods in RCM is hindered by low image quality due to noise and/or poor contrast. Use of supervised machine learning is limited by time-consuming manual labeling of RCM images.

CONCLUSIONS

RCM has great potential in the study of skin structures. The use of artificial intelligence could enable an easier, more reproducible, precise, and rigorous study of RCM images for the understanding of skin structures, skin barrier, and skin inflammation and lesions. Although several attempts have been made, further work is still needed to provide a definite gold standard and overcome issues related to image quality, limited labeled datasets, and lack of phenotype variability in available databases.

Elongated microparticles tuned for targeting hyaluronic acid delivery to specific skin strata

OBJECTIVE

Microneedle or fractional laser applications are the most common topical delivery enhancement platforms. However, these methods of drug delivery are not skin strata specific. Drug delivery approaches which could target specific stratum of the skin remains a challenge. Elongated microparticles (EMPs) have been used in enhancing drug delivery into the skin. The aim of this study was to evaluate, for the first time, elongated silica microparticles with two different length profiles to enhance delivery of hyaluronic acid into different strata of human skin.

METHODS

Two types of EMPs - long (milled EMPs) or short (etched EMPs) length ranges were characterized. A prototypical liquid formulation (Fluorescent hyaluronic acid) with and without EMP enhancement were evaluated for hyaluronic acid delivery in ex-vivo human skin. High Performance Liquid Chromatography (HPLC), Typhoon fluorescence scanning system, Laser Scanning Confocal Microscopy (LSCM) and Reflectance Confocal Microscopy (RCM) were used to validate F-HA stability, visualize fluorescein in the skin, image the depth of F-HA delivery in the skin and define EMP penetration in skin strata, respectively. Statistical analysis was conducted using GraphPad Prism 6 software (GraphPad Software Inc, USA).

RESULTS

Fluorescein-hyaluronic acid was stable and EMP enhanced skin penetration. Reflectance confocal microscopy revealed that "etched EMP" penetrated the skin to the stratum spinosum level. The vast majority (97.8%; p < 0.001) of the etched EMP did not penetrate completely through the viable epidermis and no obvious penetration into the dermis. In contrast, milled EMP showed 41-fold increase in penetration compared to the etched EMP but penetrated beyond the dermoepidermal junction.

CONCLUSION

EMPs can enhance delivery of hyaluronic acid. Using EMPs with defined length distributions, which can be tuned for a specific stratum of the skin, can achieve targeted hyaluronic acid delivery.

Towards data-driven quantification of skin ageing using reflectance confocal microscopy

INTRODUCTION

Evaluation of skin ageing is a non-standardized, subjective process, with typical measures relying coarse, qualitatively defined features. Reflectance confocal microscopy depth stacks contain indicators of both chrono-ageing and photo-ageing. We hypothesize that an ageing scale could be constructed using machine learning and image analysis, creating a data-driven quantification of skin ageing without human assessment.

METHODS

En-face sections of reflectance confocal microscopy depth stacks from the dorsal and volar forearm of 74 participants (36/18/20 training/testing/validation) were represented using a histogram of visual features learned using unsupervised clustering of small image patches. A logistic regression classifier was trained on these histograms to differentiate between stacks from 20- to 30-year-old and 50- to 70-year-old volunteers. The probabilistic output of the logistic regression was used as the fine-grained ageing score for that stack in the testing set ranging from 0 to 1. Evaluation was performed in two ways: on the test set, the AUC was collected for the binary classification problem as well as by statistical comparison of the scores for age and body site groups. Final validation was performed by assessing the accuracy of the ageing score measurement on 20 depth stacks not used for training or evaluating the classifier.

RESULTS

The classifier effectively differentiated stacks from age groups with a test set AUC of 0.908. Mean scores were significantly different when comparing age groups (mean 0.70 vs. 0.44; t = -6.62, p = 0.0000) and also when comparing stacks from dorsal and volar body sites (mean 0.64 vs. 0.53; t = 3.12, p = 0.0062). On the final validation set, 17 out of 20 depth stacks were correctly labelled.

DISCUSSION

Despite being limited to only coarse training information in the form of example stacks from two age groups, the trained classifier was still able to effectively discriminate between younger skin and older skin. Curiously, despite being only trained with chronological age, there was still evidence for measurable differences in age scores due to sun exposure-with marked differences in scores on sun-exposed dorsal sites of some volunteers compared with less sun-exposed volar sites. These results suggest that fine-grained data-driven quantification of skin ageing is achievable.

Skin strata delineation in reflectance confocal microscopy images using recurrent convolutional networks with attention

Reflectance confocal microscopy (RCM) is an effective non-invasive tool for cancer diagnosis. However, acquiring and reading RCM images requires extensive training and experience, and novice clinicians exhibit high discordance in diagnostic accuracy. Quantitative tools to standardize image acquisition could reduce both required training and diagnostic variability. To perform diagnostic analysis, clinicians collect a set of RCM mosaics (RCM images concatenated in a raster fashion to extend the field view) at 4-5 specific layers in skin, all localized in the junction between the epidermal and dermal layers (dermal-epidermal junction, DEJ), necessitating locating that junction before mosaic acquisition. In this study, we automate DEJ localization using deep recurrent convolutional neural networks to delineate skin strata in stacks of RCM images collected at consecutive depths. Success will guide to automated and quantitative mosaic acquisition thus reducing inter operator variability and bring standardization in imaging. Testing our model against an expert labeled dataset of 504 RCM stacks, we achieved [Formula: see text] classification accuracy and nine-fold reduction in the number of anatomically impossible errors compared to the previous state-of-the-art.

A Clinical Perspective on the Automated Analysis of Reflectance Confocal Microscopy in Dermatology

BACKGROUND AND OBJECTIVES

Non-invasive optical imaging has the potential to provide a diagnosis without the need for biopsy. One such technology is reflectance confocal microscopy (RCM), which uses low power, near-infrared laser light to enable real-time in vivo visualization of superficial human skin from the epidermis down to the papillary dermis. Although RCM has great potential as a diagnostic tool, there is a need for the development of reliable image analysis programs, as acquired grayscale images can be difficult and time-consuming to visually assess. The purpose of this review is to provide a clinical perspective on the current state of artificial intelligence (AI) for the analysis and diagnostic utility of RCM imaging.

STUDY DESIGN/MATERIALS AND METHODS

A systematic PubMed search was conducted with additional relevant literature obtained from reference lists.

RESULTS

Algorithms used for skin stratification, classification of pigmented lesions, and the quantification of photoaging were reviewed. Image segmentation, statistical methods, and machine learning techniques are among the most common methods used to analyze RCM image stacks. The poor visual contrast within RCM images and difficulty navigating image stacks were mediated by machine learning algorithms, which allowed the identification of specific skin layers.

CONCLUSIONS

AI analysis of RCM images has the potential to increase the clinical utility of this emerging technology. A number of different techniques have been utilized but further refinements are necessary to allow consistent accurate assessments for diagnosis. The automated detection of skin cancers requires more development, but future applications are truly boundless, and it is compelling to envision the role that AI will have in the practice of dermatology. Lasers Surg. Med. © 2020 Wiley Periodicals LLC.

Automatic Quality Assessment of Reflectance Confocal Microscopy Mosaics using Attention-Based Deep Neural Network

Skin cancers are the most common cancers with an increased incidence, and a valid, early diagnosis may significantly reduce its morbidity and mortality. Reflectance confocal microscopy (RCM) is a relatively new, non-invasive imaging technique that allows screening lesions at a cellular resolution. However, one of the main disadvantages of the RCM is frequently occurring artifacts which makes the diagnostic process more time consuming and hard to automate using e.g. end-to-end deep learning approach. A tool to automatically determine the RCM mosaic quality could be beneficial for both the lesion classification and informing the user (dermatologist) about its quality in real-time, during the examination procedure. In this work, we propose an attention-based deep network to automatically determine if a given RCM mosaic has an acceptable quality. We achieved accuracy above 87% on the test set which may considerably improve further classification results and the RCM-based examination.

Utilizing Machine Learning for Image Quality Assessment for Reflectance Confocal Microscopy

In vivo reflectance confocal microscopy (RCM) enables clinicians to examine lesions' morphological and cytological information in epidermal and dermal layers while reducing the need for biopsies. As RCM is being adopted more widely, the workflow is expanding from real-time diagnosis at the bedside to include a capture, store, and forward model with image interpretation and diagnosis occurring offsite, similar to radiology. As the patient may no longer be present at the time of image interpretation, quality assurance is key during image acquisition. Herein, we introduce a quality assurance process by means of automatically quantifying diagnostically uninformative areas within the lesional area by using RCM and coregistered dermoscopy images together. We trained and validated a pixel-level segmentation model on 117 RCM mosaics collected by international collaborators. The model delineates diagnostically uninformative areas with 82% sensitivity and 93% specificity. We further tested the model on a separate set of 372 coregistered RCM-dermoscopic image pairs and illustrate how the results of the RCM-only model can be improved via a multimodal (RCM + dermoscopy) approach, which can help quantify the uninformative regions within the lesional area. Our data suggest that machine learning-based automatic quantification offers a feasible objective quality control measure for RCM imaging.

Classification of Lentigo Maligna at Patient-Level by Means of Reflectance Confocal Microscopy Data

Reflectance confocal microscopy is an appropriate tool for the diagnosis of lentigo maligna. Compared with dermoscopy, this device can provide abundant information as a mosaic and/or a stack of images. In this particular context, the number of images per patient varied between 2 and 833 images and the objective, ultimately, is to be able to discern between benign and malignant classes. First, this paper evaluated classification at the image level, with the help of handcrafted methods derived from the literature and transfer learning methods. The transfer learning feature extraction methods outperformed the handcrafted feature extraction methods from literature, with a F 1 score value of 0.82. Secondly, this work proposed patient-level supervised methods based on image decisions and a comparison of these with multi-instance learning methods. This study achieved comparable results to those of the dermatologists, with an auc score of 0.87 for supervised patient diagnosis and an auc score of 0.88 for multi-instance learning patient diagnosis. According to these results, computer-aided diagnosis methods presented in this paper could be easily used in a clinical context to save time or confirm a diagnosis and can be oriented to detect images of interest. Also, this methodology can be used to serve future works based on multimodality.

Three-dimensional conditional random field for the dermal-epidermal junction segmentation

The segmentation of the dermal-epidermal junction (DEJ) in in vivo confocal images represents a challenging task due to uncertainty in visual labeling and complex dependencies between skin layers. We propose a method to segment the DEJ surface, which combines random forest classification with spatial regularization based on a three-dimensional conditional random field (CRF) to improve the classification robustness. The CRF regularization introduces spatial constraints consistent with skin anatomy and its biological behavior. We propose to specify the interaction potentials between pixels according to their depth and their relative position to each other to model skin biological properties. The proposed approach adds regularity to the classification by prohibiting inconsistent transitions between skin layers. As a result, it improves the sensitivity and specificity of the classification results.

Edge detector-based automatic segmentation of the skin layers and application to moisturization in high-resolution 3 Tesla magnetic resonance imaging

INTRODUCTION

Previous studies have demonstrated the feasibility to explore moisturization with quantification imaging based on T2 mapping. The aim of this study was to describe and validate the first robust automated method to segment the first layers of the skin.

MATERIALS AND METHODS

Data were picked from a previous study that included 35 healthy subjects who underwent a 3T MRI (multi spin echo calculation T2-weighted sequence) with a microscopic coil on the left heel before and one hour after moisturization. The automatic algorithm was composed of the T2 map generation, a Canny filter, a selection of boundaries, and a local regression to delimitate stratum corneum, epidermis, and dermis. An automated affine registration was applied between the exams before and after moisturization.

RESULTS

The failure rate of the algorithm was below 5%. Mean computation time was 139.12s. There was a significant and strong correlation between the automatic measurements and the manual ones for the T2 values (ρ: 0.905, P < 0.001) and for the thickness measurements (ρ: 0.8663; P < 0.001). For registration, mean of the Dice index was 0.64 [0.47; 0.80] and of the Hausdorff distance was 0.29 mm 95% CI: [0.28; 0.30].

CONCLUSION

The proposed automatic method to study the first skin layers in 3T MRI using micro-coils was robust and described T2 values and thickness measurements with a strong correlation to manual measurements. The use of an automated affine registration could also permit the generation of a mapping for a visual assessment of moisturization.

Wavelet-based statistical classification of skin images acquired with reflectance confocal microscopy

Detecting skin lentigo in reflectance confocal microscopy images is an important and challenging problem. This imaging modality has not yet been widely investigated for this problem and there are a few automatic processing techniques. They are mostly based on machine learning approaches and rely on numerous classical image features that lead to high computational costs given the very large resolution of these images. This paper presents a detection method with very low computational complexity that is able to identify the skin depth at which the lentigo can be detected. The proposed method performs multiresolution decomposition of the image obtained at each skin depth. The distribution of image pixels at a given depth can be approximated accurately by a generalized Gaussian distribution whose parameters depend on the decomposition scale, resulting in a very-low-dimension parameter space. SVM classifiers are then investigated to classify the scale parameter of this distribution allowing real-time detection of lentigo. The method is applied to 45 healthy and lentigo patients from a clinical study, where sensitivity of 81.4% and specificity of 83.3% are achieved. Our results show that lentigo is identifiable at depths between 50μm and 60μm, corresponding to the average location of the the dermoepidermal junction. This result is in agreement with the clinical practices that characterize the lentigo by assessing the disorganization of the dermoepidermal junction.

Automated Segmentation of Skin Strata in Reflectance Confocal Microscopy Depth Stacks

Reflectance confocal microscopy (RCM) is a powerful tool for in-vivo examination of a variety of skin diseases. However, current use of RCM depends on qualitative examination by a human expert to look for specific features in the different strata of the skin. Developing approaches to quantify features in RCM imagery requires an automated understanding of what anatomical strata is present in a given en-face section. This work presents an automated approach using a bag of features approach to represent en-face sections and a logistic regression classifier to classify sections into one of four classes (stratum corneum, viable epidermis, dermal-epidermal junction and papillary dermis). This approach was developed and tested using a dataset of 308 depth stacks from 54 volunteers in two age groups (20–30 and 50–70 years of age). The classification accuracy on the test set was 85.6%. The mean absolute error in determining the interface depth for each of the stratum corneum/viable epidermis, viable epidermis/dermal-epidermal junction and dermal-epidermal junction/papillary dermis interfaces were 3.1 μm, 6.0 μm and 5.5 μm respectively. The probabilities predicted by the classifier in the test set showed that the classifier learned an effective model of the anatomy of human skin.