Bronchovascular Tree-Guided Weakly Supervised Learning Method for Pulmonary Segment Segmentation

2025-05-20

AI Summary

Accurate pulmonary segment segmentation is critical for precise lung cancer localization and surgical planning, yet pixel-level annotations are prohibitively expensive. To address this, we propose an Anatomical Hierarchy-based Weakly Supervised Learning (AHSL) framework that leverages the bronchovascular tree (arteries, airways, veins) and hierarchical anatomical relationships between lobes and segments, requiring only coarse lobe- and segment-level labels for training. Our key contributions are: (1) the first anatomical hierarchy-aware supervision paradigm for weakly supervised segmentation; (2) a bronchovascular prior-guided two-stage segmentation strategy; and (3) a boundary-smoothness consistency loss coupled with anatomy-informed evaluation metrics. Evaluated on a private clinical dataset, AHSL significantly improves segmentation boundary smoothness and anatomical plausibility. Both qualitative visual assessments and quantitative results demonstrate consistent superiority over state-of-the-art weakly supervised methods.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Semi-Supervised Learning
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Pulmonary segment segmentation is crucial for cancer localization and surgical planning. However, the pixel-wise annotation of pulmonary segments is laborious, as the boundaries between segments are indistinguishable in medical images. To this end, we propose a weakly supervised learning (WSL) method, termed Anatomy-Hierarchy Supervised Learning (AHSL), which consults the precise clinical anatomical definition of pulmonary segments to perform pulmonary segment segmentation. Since pulmonary segments reside within the lobes and are determined by the bronchovascular tree, i.e., artery, airway and vein, the design of the loss function is founded on two principles. First, segment-level labels are utilized to directly supervise the output of the pulmonary segments, ensuring that they accurately encompass the appropriate bronchovascular tree. Second, lobe-level supervision indirectly oversees the pulmonary segment, ensuring their inclusion within the corresponding lobe. Besides, we introduce a two-stage segmentation strategy that incorporates bronchovascular priori information. Furthermore, a consistency loss is proposed to enhance the smoothness of segment boundaries, along with an evaluation metric designed to measure the smoothness of pulmonary segment boundaries. Visual inspection and evaluation metrics from experiments conducted on a private dataset demonstrate the effectiveness of our method.

Problem

• Weakly supervised learning for pulmonary segment segmentation
• Reducing laborious pixel-wise annotation in medical images
• Improving segmentation accuracy using bronchovascular tree guidance

Innovation

• Weakly supervised learning with anatomical hierarchy
• Two-stage segmentation using bronchovascular prior
• Consistency loss for smooth segment boundaries