C3VD-DEFCOL: A Deformable Colonoscopy Dataset with Time-Resolved 3D Ground Truth and Realistic Appearance
Researchers have introduced C3VD-DEFCOL, a new dataset designed to improve 3D reconstruction for colonoscopy procedures. This dataset provides realistic in vivo appearance alongside dense, time-resolved 3D ground truth, addressing a gap in current resources. It includes controlled deformations to simulate colon movement and peristalsis, enabling better evaluation of reconstruction algorithms and pose estimation under challenging conditions. AI
IMPACT Enables more accurate AI-driven 3D reconstruction for medical procedures, potentially improving diagnostic capabilities.