Search papers, labs, and topics across Lattice.
Equal supervision
4
0
7
RadGrounder achieves competitive performance in radiology VQA while enabling spatial grounding without compromising language quality.
Current LLMs falter in delivering reliable medical assistance, exposing a critical gap in their ability to coordinate knowledge, communication, and EHR interactions.
Radiologists routinely pinpoint specific CT scan slices in their reports, and now AI can learn to "read where to look" too, grounding language in 3D volumes with unprecedented accuracy.
Medical VLMs get a calibration boost without training or labels: LATA sharpens predictions by smoothing over a k-NN graph, shrinking prediction sets and balancing class coverage.
