Work

LENAAFMIS: Learning from Explainable and Navigated Attention for Annotation-Free Medical Image Segmentation

Year: 2026
Type: software
Abstract: Code for LENAAFMIS, an annotation-free medical image segmentation framework that converts explainability maps of a differential vision-language classifier into prompts for a promptable segmentation model, with confidence-driven iterative refinement. Version used in the Communications AI & Computing article.
Language: English
Open Access status: green
OpenAlex ID: W7214734366
Cites:
Cited by:
Related to: 10