Artificial intelligence in dermatovenereology: a review of mobile apps with AI functionality

Abstract

Aim. To conduct a systematic review of artificial intelligence (AI)-powered mobile apps in dermatovenereology, assess their number, target audience, functionality, level of scientific validation, regulatory status, and data security. Materials and methods. We searched the Google Play Store and Apple App Store for mobile apps based on AI models (43 apps were selected from 909 records), and Google Scholar for publications from 2020–2025 (seven studies on the use of convolutional neural networks in mobile apps were selected from 209 records). Results. Of the 43 apps, 74.4% were patient-focused, 14.0% were physician-focused, and 11.6% were both. The main tasks were: skin cancer screening (32.6%), skin disease diagnosis (27.9%), mole monitoring (14.0%), and acne (11.6%). Only 7 apps (16.3%) had peer-reviewed publications. The Russian apps Derma Onko Check and Melanoma Check stand out for their validated datasets, publications, and full on-device data processing. Conclusion. Most mobile AI apps in dermatovenereology do not meet the minimum requirements for scientific validation, regulatory certification, and ethical transparency. Safe use requires independent publications, disclosure of training data characteristics, local image processing, and clear regulatory status.

Ähnliche Arbeiten