Researchers have introduced BifDet, a new dataset designed for detecting 3D airway bifurcations in CT scans. This dataset addresses a significant gap in resources for analyzing lung physiology and disease mechanisms. BifDet includes annotated CT scans from the ATM22 cohort, with bounding boxes for parent and daughter airway branches. The paper also demonstrates the dataset's utility by fine-tuning and evaluating RetinaNet and DETR models for bifurcation detection, providing baseline results for future research. AI
Summary written by gemini-2.5-flash-lite from 2 sources. How we write summaries →
IMPACT Provides a specialized dataset for advancing AI-driven analysis of respiratory diseases through airway tree modeling.
RANK_REASON The cluster describes the release of a new dataset and associated research paper on arXiv.