Communications of the ACM
ACM Opinion
Medical AI Is as Much Social as It Is Technological
With limited or no information on how an AI system analyzes data or comes to a conclusion, there is a lack of ‘explainability,’ which can be a serious problem in healthcare settings
Credit: Getty Images
Despite the promise of medical artificial intelligence (AI) applications, their acceptance in real-world clinical settings is low, with lack of transparency and trust being barriers that need to be overcome.
Thinking of AI applications for clinical use as an intricate combination of scientific, technological, and social inputs will ultimately make research and development in this area more robust, trustworthy, and successful.
From Nature Machine Intelligence
View Full Article (May Require Paid Registration)
No entries found