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.

Background Elderly patients in senior communities faced high barriers to care during the COVID‐19 pandemic, including increased vulnerability to COVID‐19, long quarantines for clinic visits, and difficulties with telemedicine adoption. Objective To pilot a new model of dermatologic care to overcome barriers for senior living communities during the COVID‐19 pandemic and assess patient satisfaction. Methods From 16 November 2020 to 9 July 2021, this quality improvement programme combined in‐residence full body imaging with real‐time outlier lesion identification and virtual teledermatology. Residents from the Sequoias Portola Valley Senior Living Retirement Community (Portola Valley, California) voluntarily enroled in the Stanford Skin Scan Programme. Non‐physician clinical staff with a recent negative COVID‐19 test travelled on‐site to obtain in‐residence full body photographs using a mobile app‐based system on an iPad called SkinIO that leverages deep learning to analyse patient images and suggest suspicious, outlier lesions for dermoscopic photos. A single dermatologist reviewed photographs with the patient and provided recommendations via a video visit. Objective measures included follow‐up course and number of skin cancers detected. Subjective findings were obtained through patient experience surveys. Results Twenty‐seven individuals participated, three skin cancers were identified, with 11 individuals scheduled for a follow up in‐person visit and four individuals starting home treatment. Overall, 88% of patients were satisfied with the Skin Scan programme, with 77% likely to recommend the programme to others. 92% of patients agreed that the Skin Scan photographs were representative of their skin. In the context of the COVID‐19 pandemic, 100% of patients felt the process was safer or comparable to an in‐person visit. Despite overall appreciation for the programme, 31% of patients reported that they would prefer to see dermatologist in‐person after the pandemic. Conclusions This programme offers a framework for how a hybrid skin scan programme may provide high utility for individuals with barriers to accessing in‐person clinics.

J.M.B. and M.K. contributed equally to this work.

ImportanceTelemedicine use accelerated during the COVID-19 pandemic, and skin conditions were a common use case. However, many images submitted may be of insufficient quality for making a clinical determination.ObjectiveTo determine whether an artificial intelligence (AI) decision support tool, a machine learning algorithm, could improve the quality of images submitted for telemedicine by providing real-time feedback and explanations to patients.Design, Setting, and ParticipantsThis quality improvement study with an AI performance component and single-arm clinical pilot study component was conducted from March 2020 to October 2021. After training, the AI decision support tool was tested on 357 retrospectively collected telemedicine images from Stanford telemedicine from March 2020 to June 2021. Subsequently, a single-arm clinical pilot study was conducted to assess feasibility with 98 patients in the Stanford Department of Dermatology across 2 clinical sites from July 2021 to October 2021. For the clinical pilot study, inclusion criteria for patients included being adults (aged ≥18 years), presenting to clinic for a skin condition, and being able to photograph their own skin with a smartphone.InterventionsDuring the clinical pilot study, patients were given a handheld smartphone device with a machine learning algorithm interface loaded and were asked to take images of any lesions of concern. Patients were able to review and retake photos prior to submitting, so each submitted photo met the patient’s assumed standard of clinical acceptability. A machine learning algorithm then gave the patient feedback on whether the image was acceptable. If the image was rejected, the patient was provided a reason by the AI decision support tool and allowed to retake the photos.Main Outcomes and MeasuresThe main outcome of the retrospective image analysis was the receiver operator curve area under the curve (ROC-AUC). The main outcome of the clinical pilot study was the image quality difference between the baseline images and the images approved by AI decision support.ResultsOf the 98 patients included, the mean (SD) age was 49.8 (17.6) years, and 50 (51%) of the patients were male. On retrospective telemedicine images, the machine learning algorithm effectively identified poor-quality images (ROC-AUC of 0.78) and the reason for poor quality (blurry ROC-AUC of 0.84; lighting issues ROC-AUC of 0.70). The performance was consistent across age and sex. In the clinical pilot study, patient use of the machine learning algorithm was associated with improved image quality. An AI algorithm was associated with reduction in the number of patients with a poor-quality image by 68.0%.Conclusions and RelevanceIn this quality improvement study, patients use of the AI decision support with a machine learning algorithm was associated with improved quality of skin disease photographs submitted for telemedicine use.

No abstract