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

Artificial intelligence and skin cancer

Artificial intelligence is poised to rapidly reshape many fields, including that of skin cancer screening and diagnosis, both as a disruptive and assistive technology. Together with the collection and availability of large medical data sets, artificial intelligence will become a powerful tool that can be leveraged by physicians in their diagnoses and treatment plans for patients. This comprehensive review focuses on current progress toward AI applications for patients, primary care providers, dermatologists, and dermatopathologists, explores the diverse applications of image and molecular processing for skin cancer, and highlights AI's potential for patient self-screening and improving diagnostic accuracy for non-dermatologists. We additionally delve into the challenges and barriers to clinical implementation, paths forward for implementation and areas of active research.

A New Assessment Method of Vitiligo by Combination of Dermoscopy and Reflectance Confocal Microscopy

Purpose

The aim is to investigate the application value of dermoscopy combined with reflectance confocal microscopy (RCM) in assessing vitiligo disease activity and treatment response.

Patients and Methods

We enrolled 279 patients with vitiligo and evaluated the disease activity by Vitiligo Disease Activity (VIDA) score, dermoscopy, RCM and dermoscopy combined with RCM respectively. The sensitivity and specificity of different assessment techniques were compared with VIDA score by the differences and consistency. The different characteristics of dermoscopy and RCM with different treatment responses were also analyzed.

Results

The results showed that the sensitivity and specificity of dermoscopy combined RCM were higher than RCM or dermoscopy alone (P values less than 0.05). In the repigmentation process, leukotrichia, pigment network absent and perilesional hyperpigmentation under dermoscopy at the baseline suggested a poor treatment response, while the incompletely disappearing pigment rings under RCM and perifollicular hyperpigmentation under dermoscopy indicated a good treatment response. We also found the proportion of patients with telangiectasia, increased pigment at the lesions and around the hair follicles was significantly higher in the good treatment response group than that in the poor one by dermoscopy (χ2 = 4.423, 32.471, 4.348, P = 0.035 0.000, 0.037) and by RCM the proportion of patients with both increased pigment granules and dendritic melanocytes in the good treatment response group was higher than that in the poor one (χ2 = 38.215, 5.283, P = 0.000, 0.022, respectively).

Conclusion

With the higher sensitivity and specificity than dermoscopy or RCM alone, a combination of dermoscopy and RCM may be a new more accurate measure to assess the vitiligo disease activity and the treatment response.

The role of in vivo reflectance confocal microscopy for the management of acne: A systematic review

Acne diagnosis, severity assessment and treatment follow-up rely primarily on clinical examination. In vivo reflectance confocal microscopy (RCM) provides non-invasively, real-time images of skin lesions with a level of detail close to histopathology. This systematic literature review aims to provide an overview of RCM utility in acne and a summary of specific features with clinical application that may increase objectivity in evaluating this condition. We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for reporting our results. We systematically searched three databases: PubMed, Clarivate and Google Scholar (January 2022). All included studies used RCM to investigate acne in human patients and reported the investigated skin area and type (acne lesions or clinically uninvolved skin), the substance used in the case of treatment. Our search identified 2184 records in the three databases investigated. After duplicate removal, 1608 records were screened, 35 were selected for full-text assessment, and 14 were included in this review. We used the QUADAS-2 tool to evaluate the risk of bias and applicability concerns. RCM was selected as the index test and clinical examination as the reference standard. The total number of patients from all studies was 291, with 216 acne patients and 60 healthy participants aged between 13 and 45 years. The 14 considered studies analysed 456 follicles from healthy participants, 1445 follicles from uninvolved skin in acne patients and 1472 acne lesions. Consistent RCM findings concerning follicles of acne patients reported across studies were increased follicular infundibulum size, thick, bright border, intrafollicular content and inflammation. Our analysis indicates that RCM is a promising tool for acne evaluation. Nevertheless, standardization, a unified terminology, consistent research methods and unitary reporting of RCM findings are necessary. PROSPERO registration number CRD42021266547.

Artificial intelligence-assisted dermatology diagnosis: From unimodal to multimodal

Artificial Intelligence (AI) is progressively permeating medicine, notably in the realm of assisted diagnosis. However, the traditional unimodal AI models, reliant on large volumes of accurately labeled data and single data type usage, prove insufficient to assist dermatological diagnosis. Augmenting these models with text data from patient narratives, laboratory reports, and image data from skin lesions, dermoscopy, and pathologies could significantly enhance their diagnostic capacity. Large-scale pre-training multimodal models offer a promising solution, exploiting the burgeoning reservoir of clinical data and amalgamating various data types. This paper delves into unimodal models' methodologies, applications, and shortcomings while exploring how multimodal models can enhance accuracy and reliability. Furthermore, integrating cutting-edge technologies like federated learning and multi-party privacy computing with AI can substantially mitigate patient privacy concerns in dermatological datasets and further fosters a move towards high-precision self-diagnosis. Diagnostic systems underpinned by large-scale pre-training multimodal models can facilitate dermatology physicians in formulating effective diagnostic and treatment strategies and herald a transformative era in healthcare.

Adaptability of melanocytes post ultraviolet stimulation in patients with melasma

BACKGROUND

Dynamic in vivo changes in melanin in melasma lesions after exposure to ultraviolet (UV) irradiation have not been described.

OBJECTIVES

To determine whether melasma lesions and nearby perilesions demonstrated different adaptive responses to UV irradiation and whether the tanning responses were different among different locations on face.

METHODS

We collected sequential images from real-time cellular resolution full-field optical coherence tomography (CRFF-OCT) at melasma lesions and perilesions among 20 Asian patients. Quantitative and layer distribution analyses for melanin were performed using a computer-aided detection (CADe) system that utilizes spatial compounding-based denoising convolutional neural networks.

RESULTS

The detected melanin (D) is melanin with a diameter >0.5 µm, among which confetti melanin (C) has a diameter of >3.3 µm and corresponds to a melanosome-rich package. The calculated C/D ratio is proportional to active melanin transportation. Before UV exposure, melasma lesions had more detected melanin (p = 0.0271), confetti melanin (p = 0.0163), and increased C/D ratio (p = 0.0152) in the basal layer compared to those of perilesions. After exposure to UV irradiation, perilesions have both increased confetti melanin (p = 0.0452) and the C/D ratio (p = 0.0369) in basal layer, and this effect was most prominent in right cheek (p = 0.030). There were however no significant differences in the detected, confetti, or granular melanin areas before and after exposure to UV irradiation in melasma lesions in all the skin layers.

CONCLUSIONS

Hyperactive melanocytes with a higher baseline C/D ratio were noted in the melasma lesions. They were "fixed" on the plateau and were not responsive to UV irradiation regardless of the location on face. Perilesions retained adaptability with a dynamic response to UV irradiation, in which more confetti melanin was shed, mainly in the basal layer. Therefore, aggravating effect of UV on melasma was mainly due to UV-responsive perilesions rather than lesions.

Artificial Intelligence-Based Approaches to Reflectance Confocal Microscopy Image Analysis in Dermatology

Reflectance confocal microscopy (RCM) is a non-invasive imaging method designed to identify various skin diseases. Confocal based diagnosis may be subjective due to the learning curve of the method, the scarcity of training programs available for RCM, and the lack of clearly defined diagnostic criteria for all skin conditions. Given that in vivo RCM is becoming more widely used in dermatology, numerous deep learning technologies have been developed in recent years to provide a more objective approach to RCM image analysis. Machine learning-based algorithms are used in RCM image quality assessment to reduce the number of artifacts the operator has to view, shorten evaluation times, and decrease the number of patient visits to the clinic. However, the current visual method for identifying the dermal-epidermal junction (DEJ) in RCM images is subjective, and there is a lot of variation. The delineation of DEJ on RCM images could be automated through artificial intelligence, saving time and assisting novice RCM users in studying the key DEJ morphological structure. The purpose of this paper is to supply a current summary of machine learning and artificial intelligence’s impact on the quality control of RCM images, key morphological structures identification, and detection of different skin lesion types on static RCM images.