Three stage method improves colon scans from 3D CT images
Preserving Anatomical Continuity: Three-Stage Pipeline for Colon Segmentation in 3D Abdominal CT Scans
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Segmenting the colon accurately in CT scans is hard because its shape is complex and can appear disconnected in images. The authors created a three-step process to fix this problem by first using AI to predict the colon areas, then reconnecting any broken parts, and finally refining the shape for better continuity. This approach helps keep the colon’s structure intact and makes the segmentation more reliable for doctors and researchers. They tested the method on different datasets and found it improved how well the colon’s shape is captured without losing accuracy.
What this means in practice
- •For medical imaging teams: Provide more structurally accurate colon segmentations to support colorectal disease diagnosis and treatment planning from abdominal CT scans.
- •For clinical software developers: Incorporate topology-preserving segmentation into diagnostic tools to improve reliability and consistency of colon imaging results for radiologists.$Commercial implications: Makes topology-aware colon segmentation possible for clinical imaging software products, enhancing diagnostic capabilities.
Authors
Deshan Kalupahana, Sonit Singh, Praveen Ravindran, Arcot Sowmya
Abstract
Accurate colon segmentation from CT images is essential for colorectal disease analysis, yet deep learning based methods often produce disconnected predictions due to complex anatomy. This study introduces a three-stage, topology-preserving segmentation pipeline to address this issue. The first stage performs initial deep learning-based segmentation, followed by centreline bridging to reconnect disjoint regions and a reconstruction stage to refine continuity. Evaluations on TotalSegmentator and RAOS datasets using overlap, distance and topology-based metrics demonstrate improved structural consistency while maintaining segmentation accuracy. The proposed method enhances topological integrity, enabling more reliable colon segmentation for clinical and research applications.