Canvas Dx

This study explores how the FDA’s predetermined change control plan (PCCP) can enhance the performance of an AI-based autism diagnostic device through iterative learning. By optimizing decision thresholds using a dataset of 722 children, the study improved the device’s ability to accurately detect or rule out autism without altering its intended use. The results highlight the potential of adaptive regulatory mechanisms to improve medical devices’ real-world performance, particularly in addressing the growing demand for autism evaluations in the United States.