Study Raises Concerns About Reliability of AI Models for Personalized Medicine
-
Mathematical models used for personalized medicine are effective within specific clinical trials but fail to generalize across different trials.
-
The study raises concerns about applying AI and machine learning to personalized medicine, especially for variable conditions like schizophrenia.
-
Algorithms often "overfit" to initial trial data and don't work on new data from different trials.
-
More data sharing and inclusion of additional variables could improve reliability of AI algorithms for medical treatments.
-
Since models don't generalize across trials, there are limitations in current personalized medicine, underscoring a gap between potential and real-world application.