Comprehensive Summary
This article studies atopic dermatitis, a skin condition characterized by pruritus and skin lesions, and the most common form of eczema. Atopic dermatitis affects approximately 20% of all pediatric patients. Current emerging artificial intelligence tools, which utilize convolutional neural networks in assessing skin lesions, aim to diagnose and manage atopic dermatitis. Yip et al. analyzed 3122 papers on development of such AI tools. The current state of AI tools in atopic dermatitis are actively advancing, but require more development before they can reliably be used in diagnosis and condition monitoring.
Outcomes and Implications
As atopic dermatitis affects a large proportion of pediatric patients, AI tools have a large amount of potential in identifying and treating the condition. The authors note that AI tools require much more development and would need to shift their focus to patient care and clinical decision making to be deemed reliable in a clinical setting. Yip et al. suggest collaboration between experts in academic, clinical, and industry settings in order to develop the AI models further before beginning integration into clinical settings.