Skin cancer, the most common human malignancy, is primarily diagnosed visually, beginning with an initial clinical screening and followed potentially by dermoscopic analysis, a biopsy and histopathological examination. Automated classification of skin lesions using images is a challenging task owing to the fine-grained variability in the appearance of skin lesions. Deep convolutional neural networks (CNNs) show potential for general and highly variable tasks across many fine-grained object categories. Here we demonstrate classification of skin lesions using a single CNN, trained end-to-end from images directly, using only pixels and disease labels as inputs. We train a CNN using a dataset of 129,450 clinical images-two orders of magnitude larger than previous datasets-consisting of 2,032 different diseases. We test its performance against 21 board-certified dermatologists on biopsy-proven clinical images with two critical binary classification use cases: keratinocyte carcinomas versus benign seborrheic keratoses; and malignant melanomas versus benign nevi. The first case represents the identification of the most common cancers, the second represents the identification of the deadliest skin cancer. The CNN achieves performance on par with all tested experts across both tasks, demonstrating an artificial intelligence capable of classifying skin cancer with a level of competence comparable to dermatologists. Outfitted with deep neural networks, mobile devices can potentially extend the reach of dermatologists outside of the clinic. It is projected that 6.3 billion smartphone subscriptions will exist by the year 2021 (ref. 13) and can therefore potentially provide low-cost universal access to vital diagnostic care.

Background: PRAME (PReferentially expressed Antigen in MElanoma) has shown utility in distinguishing melanoma from benign melanocytic lesions, but knowledge of its expression pattern in intermediate melanocytic and spitzoid proliferations is limited. Methods: Immunohistochemical expression of PRAME was examined in 112 melanocytic proliferations with intermediate histopathologic or spitzoid features. Results: Any intensity of nuclear PRAME staining in at least 60% of lesional melanocytes was determined as the best threshold for diffuse staining in this cohort. Nearly all non-spitzoid melanomas (23/24; 95.8%) demonstrated diffuse PRAME expression. PRAME was completely negative in 95.6% (43/45) of mitotically-active nevi, traumatized nevi, nevi with persistent/recurrent features, and dysplastic nevi. Most Spitz nevi (15/20) and atypical Spitz tumors (10/13) entirely lacked PRAME expression. One Spitz nevus, one atypical Spitz tumor, and one spitzoid melanoma (1/2) demonstrated diffuse PRAME expression. Conclusions: Although diffuse PRAME expression is generally limited to malignant melanoma, benign Spitz nevi and atypical Spitz tumors can infrequently express diffuse PRAME. PRAME immunohistochemistry can be useful in the evaluation of atypical melanocytic proliferations with intermediate histopathologic features but should be interpreted with caution in the setting of spitzoid neoplasms.

These findings expand the literature of immune-related toxicities of PD-L1 and PD-1 blockade to include lichenoid dermatitis and lichenoid mucositis. Of note, these cutaneous side effects were amenable to topical treatment, without the need for medication dose reduction or discontinuation.

No abstract

An estimated 3 billion people lack access to dermatological care globally. Artificial intelligence (AI) may aid in triaging skin diseases and identifying malignancies. However, most AI models have not been assessed on images of diverse skin tones or uncommon diseases. Thus, we created the Diverse Dermatology Images (DDI) dataset—the first publicly available, expertly curated, and pathologically confirmed image dataset with diverse skin tones. We show that state-of-the-art dermatology AI models exhibit substantial limitations on the DDI dataset, particularly on dark skin tones and uncommon diseases. We find that dermatologists, who often label AI datasets, also perform worse on images of dark skin tones and uncommon diseases. Fine-tuning AI models on the DDI images closes the performance gap between light and dark skin tones. These findings identify important weaknesses and biases in dermatology AI that should be addressed for reliable application to diverse patients and diseases.