Clinical artificial intelligence (AI) systems are susceptible to performance degradation due to data shifts, which can lead to erroneous predictions and potential patient harm.
Proactively detecting and mitigating these shifts is crucial for maintaining AI effectiveness and safety in clinical practice.
This prognostic study was conducted using electronic health record data for admissions to general internal medicine wards of 7 large hospitals in Toronto, Canada, between January 1, 2010, to August 31, 2020.
Significant data shifts due to changes in demographics, hospital types, admission sources, and critical laboratory assays were detected using a label-agnostic monitoring pipeline.
Transfer learning and drift-triggered continual learning strategies mitigated these shifts and significantly improved model performance.
The findings suggest that a proactive, label-agnostic monitoring pipeline incorporating transfer and continual learning can detect and mitigate harmful data shifts in Toronto's general internal medicine population, ensuring robust and equitable clinical AI deployment.
【MORE】