Learning Topology-Aware Implicit Field for Unified Pulmonary Tree Modeling with Incomplete Topological Supervision
Researchers have developed TopoField, a novel framework designed to address topological incompleteness in pulmonary trees extracted from CT scans. This implicit modeling approach treats topology repair as a primary task, enabling unified inference for anatomical labeling and lung segment reconstruction. TopoField demonstrates significant improvements in topological completeness and accuracy, even when applied to incomplete outputs from external segmentation models, offering a computationally efficient solution for clinical applications. AI