A review recently published in JAMA Dermatology analyzed the current state of artificial intelligence (AI) dermatology mobile applications (apps), looking specifically at their purpose, supporting evidence, regulatory status, clinician input, data privacy measures, and use of image data.
Researchers initially identified 909 apps for this study. However, 41 apps were included in this review following the removal of those that did not fit the relevant criteria.
Results
Concerning target audience, 32 of the apps (78%) were directed towards patients, four (9.76%) towards clinicians, and 5 (12.2%) towards both patient and clinician.
Fourteen (34.1%) of the apps were for the purpose of skin cancer detection, 13 (31.7%) for the diagnosis and/or identification of skin and/or hair conditions, and 7 (17.1%) for mole tracking. Some of the other uses included tracking skin conditions, diagnosing, treating, and monitoring acne, managing atopic dermatitis, and sun protection.
A total of 14 apps (34.1%) were U.S.-based. However, only two of these provided a disclaimer revealing the lack of U.S. Food and Drug Administration (FDA) approval. Another 14 (34.1%) were based in Europe, and similarly, only two of these included a Conformité Européenne (CE) Mark to reveal any regulatory health and safety approval information from the European Union.
Ten (24.4%) of the apps included in this review claimed to have diagnostic capabilities. Researchers noted, however, that none of these apps provided any scientific publications as supporting evidence. Furthermore, two of them also lacked any warnings to patients about any potential inaccuracies.
Researchers also noted that 24 apps (58.5%) lacked any information on training and testing datasets. Furthermore, those that did have such information offered very vague and generalized descriptions. They also noted that 21 apps (51.2%) did not provide any algorithm details. Out of these 41 apps, only 16 (39.0%) of them stated that they had sought dermatologist’s input for their development.
As it concerns the use of images, 12 (29.3%) of the 41 apps indicated that they did not store user-submitted images. A total of 16 (39.0%), however, stated that they did, 12 of which reported using secure cloud servers. Twenty apps (48.8%) reported that the images would be used for analysis to provide user results, and 12 apps (29.3%) used it for research and further app development. A total of 19 apps (46.3%) did not provide any details on this.
According to the authors, “The findings of this scoping review highlight that dermatology AI apps lack transparency about the effectiveness of the AI models, data used for development, and how user images are used. This raises concerns about biases, inappropriate recommendations, and user privacy. The absence of regulatory approval and limited clinician involvement, particularly from dermatologists, further compounds these concerns.”
They conclude that, “although AI dermatology mobile apps hold promise for improving access to care and patient outcomes, in their current state, they may pose harm due to potential risks, lack of consistent validation, and misleading user communication. Addressing challenges in efficacy, safety, and transparency through effective regulation, validation, and standardized evaluation criteria is essential to harness the benefits of these apps while minimizing risks.”
Sources:
- Wongvibulsin S, Yan MJ, Pahalyants V, Murphy W, Daneshjou R, Rotemberg V. Current State of Dermatology Mobile Applications With Artificial Intelligence Features. JAMA Dermatol. Published online March 07, 2024. doi:10.1001/jamadermatol.2024.0468
- AI Dermatology Mobile Apps Have Critical Efficacy, Safety Gaps, Review Says
