Search papers, labs, and topics across Lattice.
Affiliation:
5
0
6
LUX achieves unprecedented alignment between generated captions and localized pathological evidence, drastically reducing clinical hallucinations in endoscopic analysis.
Visibility-aware self-supervision enables robust 3D-2D liver registration even under extreme occlusion, achieving a Dice score of 92.6%.
Multimodal OOD detection gets a serious upgrade: a dual-branch approach boosts performance by nearly 25% on endoscopic images.
Finally, a comprehensive, multi-center dataset with dual scoring metrics and expert-generated captions is available to advance clinically meaningful multimodal algorithms for ulcerative colitis scoring in endoscopy.
Deep learning can spot polyps in colonoscopies better when it watches the video, not just single frames.
