Dermoscopic features of neoplasms in skin of color: A review

From data to diagnosis: skin cancer image datasets for artificial intelligence

Artificial intelligence (AI) solutions for skin cancer diagnosis continue to gain momentum, edging closer towards broad clinical use. These AI models, particularly deep-learning architectures, require large digital image datasets for development. This review provides an overview of the datasets used to develop AI algorithms and highlights the importance of dataset transparency for the evaluation of algorithm generalizability across varying populations and settings. Current challenges for curation of clinically valuable datasets are detailed, which include dataset shifts arising from demographic variations and differences in data collection methodologies, along with inconsistencies in labelling. These shifts can lead to differential algorithm performance, compromise of clinical utility, and the propagation of discriminatory biases when developed algorithms are implemented in mismatched populations. Limited representation of rare skin cancers and minoritized groups in existing datasets are highlighted, which can further skew algorithm performance. Strategies to address these challenges are presented, which include improving transparency, representation and interoperability. Federated learning and generative methods, which may improve dataset size and diversity without compromising privacy, are also examined. Lastly, we discuss model-level techniques that may address biases entrained through the use of datasets derived from routine clinical care. As the role of AI in skin cancer diagnosis becomes more prominent, ensuring the robustness of underlying datasets is increasingly important.

Dermoscopy Use in Africa: Determinants and Challenges

INTRODUCTION

Dermoscopy has evolved over the years beyond distinguishing benign pigmented lesions from melanoma to diagnosing virtually all diseases in dermatology. Overwhelming evidence demonstrates its utility in improving diagnostic accuracy, reducing unnecessary biopsies and lesion monitoring. Dermoscopy is widely used in Western nations, hence most descriptions of lesions in literature are predominantly on Fitzpatrick skin types I-III. Current evidence shows that there are unique dermoscopic features in the dark skin as a result of pigment and pathological reactions. Nationwide surveys and reports have been conducted across several continents to highlight prevalence and factors influencing dermoscopy use with the hope of maximizing its apparent benefits. There are currently no such reports from Africa.

OBJECTIVES

To evaluate dermoscopy use and its determinants among dermatologists in Africa.

METHODS

A cross-sectional study. Online forms were e-mailed to individual practicing dermatologists and members of the African Society of Dermatologists and Venereologists.

RESULTS

There were 196 respondents from 24 African countries. Half of them used dermoscopy. Training, practice settings and location, provision of dermatoscopes by institutions and knowledge of criteria were notable significant determinants. Multiple training exposures, knowledge of criteria, availability of dermatoscopes, use of both hand-held and videodermatoscopes, average number of patients seen per day, and a positive outlook towards dermoscopy were significant determinants of frequency of use. Leading impediments were lack of training and inadequate dermatoscopes in practice.

CONCLUSIONS

Dermoscopy use in Africa is relatively low. Incorporating dermoscopy training into the curriculum with provision of dermatoscopes by training institutions will promote wider usage.

Dermoscopy of Cutaneous Neoplasms in Skin of Color - A Systematic review by the International Dermoscopy Society "Imaging in Skin of Color" Task Force

Over the last few decades, dermoscopy has been showed to facilitate the non-invasive diagnosis of both benign and malignant skin tumors, yet literature data mainly comes from studies on light photo-types. However, there is growing evidence that skin neoplasms may benefit from dermoscopic assessment even for skin of color. This systematic literature review evaluated published data in dark-skinned patients (dermoscopic features, used setting, pathological correlation, and level of evidence of studies), also providing a standardized and homogeneous terminology for reported dermoscopic findings. A total of 20 articles describing 46 different tumors (four melanocytic neoplasms, eight keratinocytic tumors, 15 adnexal cutaneous neoplasms, seven vascular tumors, four connective tissue tumors, and eight cystic neoplasms/others) for a total of 1724 instances were included in the analysis. Most of them showed a level of evidence of V (12 single case reports and six case series), with only two studies featuring a level of evidence of IV (case-control analysis). Additionally, this review also underlined that some neoplasms and phototypes are underrepresented in published analyses as they included only small samples and mainly certain tones of "dark skin" spectrum (especially phototype IV). Therefore, further studies considering such limitations are required for a better characterization.

Melanoma in skin of color: Part I. Epidemiology and clinical presentation

Although there is a higher incidence of melanoma among non-Hispanic White individuals, melanoma is diagnosed at more advanced stages and associated with worse survival rates among individuals with skin of color (SOC). The proportions of melanoma subtypes differ across racial groups, with acral lentiginous melanoma and mucosal melanoma representing higher proportions of melanoma diagnoses in individuals with SOC compared to White individuals. The recognition of distinct differences in anatomic locations and dermatoscopic patterns may facilitate the appropriate differentiation of physiologic from pathologic pigmentation. The first article of this continuing medical education series will focus on the epidemiology and clinical presentation of melanoma in individuals with SOC, with the aim of improving early diagnoses and clinical outcomes.

International Dermoscopy Society (IDS) Criteria for Skin Tumors: Validation for Skin of Color Through a Delphi Expert Consensus by the "Imaging in Skin of Color" IDS Task Force

INTRODUCTION

A structured set of eight basic dermoscopic parameters (lines, clods, dots, circles, pseudopods, structureless, else, and vessels) including a total of 77 variables with corresponding descriptive and metaphoric vocabulary has been released for evaluation of skin tumors by the International Dermoscopy Society (IDS).

OBJECTIVES

To validate the aforementioned criteria for the use in darker phototypes (phototypes IV-VI) via an expert consensus.

METHODS

The two-round "Delphi method" was adopted, with an iterative process including two rounds of email questionnaires. Potential panelists were asked to take part in the procedure via email on the basis of their expertise in the dermoscopy of skin tumors in dark phototypes.

RESULTS

A total of 17 participants were involved. All the original variables of the eight basic parameters reached agreement during the first round, except for "pink small clods" ("milky red globules") and "structureless pink zone" ("milky red areas"). Moreover, during the first round, panelists proposed a change of three existing items and the introduction of four new items, i.e., "black, small clods" ("black globules"), "follicular plugs", "erosions/ulcerations", and "white color around vessels" ("perivascular white halo"). All such proposals achieved agreement, thus being included in the final list, for a total of 79 items. There was consistency between the descriptive and metaphoric approaches in terms of scoring.

CONCLUSIONS

Albeit most of the original items were considered applicable even for skin of color, there are some points of differences that physicians need to know. No significant preference was found between descriptive and metaphoric terminology among panelists.

Clinical and dermoscopic findings of benign longitudinal melanonychia due to melanocytic activation differ by skin type and predict likelihood of nail matrix biopsy

BACKGROUND

Longitudinal melanonychia (LM) is a common dermatologic finding in clinical practice with a broad differential diagnosis. Melanocytic activation is the most common LM etiology.

OBJECTIVE

To investigate clinical and dermoscopic differences of benign LM based on Fitzpatrick skin type and in biopsied vs. non-biopsied patients.

METHODS

A 10-year retrospective cohort of 248 benign LM cases at Weill Cornell Dermatology was identified and analyzed.

RESULTS

Darker-skinned vs. lighter-skinned patients had higher band width percentage (p=0.0125), lower band brightness (p<0.001), more band changes (p=0.0071), and received more biopsies (p=0.032). Biopsied (n=47) vs. non-biopsied patients (n=201) had less multidigit band involvement (p=0.0008), higher band width percentage (p=0.0213), lower band brightness (p=0.0003), and more band changes (p<0.0001). Darker skin types more often had brown vs. grey coloration on dermoscopy (p=0.0232). Mean band width percentage for all biopsied patients was 30.81% (range: 5.80%-100%).

LIMITATIONS

Single-center retrospective design. Subungual melanoma and other benign LM etiologies were not analyzed. Only 18.95% of patients received a biopsy.

CONCLUSION

Darker vs. lighter skin types more often present with darker and wider bands, brown vs. grey coloration on dermoscopy, and receive more biopsies. Multi-institutional studies on LM are needed to determine nail matrix biopsy criteria in different skin types.