Papers

A Label-Efficient On-Device QC Agent for MRI Segmentation under Domain Shift

Under domain shift a segmenter can miss a tumor confidently, so QC built on uncertainty inherits a blind spot. This agent instead learns a failure detector over a compact 29-dimensional signature of the segmenter's output, flags likely failures for clinician review, and recalibrates online from the cases that come back reviewed — without ever touching the segmenter's weights. On held-out BraTS-Africa it reaches AUROC 0.91, and reviewing the top-ranked 20% of slices surfaces 48% of all major failures; the numpy-only runtime scores a slice in 88 µs on a 15 W Jetson Orin Nano with model state under 200 KB.