SegKAN: High-Resolution Medical Image Segmentation with Long-Distance Dependencies

2024-12-28

AI Summary

To address severe vessel discontinuity and strong noise interference—leading to poor structural integrity in hepatic vascular CT image segmentation—this paper proposes SegKAN. Methodologically: (1) a convolution-enhanced embedding module is designed to effectively suppress noise and improve training stability; (2) for the first time, spatial relationships among ViT image patches are explicitly reformulated as temporal dependencies, thereby strengthening long-range contextual modeling. The model integrates the enhanced embedding, temporally structured patch representation, and a customized Transformer variant architecture. Evaluated on a hepatic vascular CT dataset, SegKAN achieves a Dice coefficient 1.78% higher than the state-of-the-art, demonstrating significantly improved segmentation completeness and robustness—particularly for slender, high-resolution vascular structures.

Technology Category

• Computer Vision: Segmentation
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Machine Learning: Structured Learning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Hepatic vessels in computed tomography scans often suffer from image fragmentation and noise interference, making it difficult to maintain vessel integrity and posing significant challenges for vessel segmentation. To address this issue, we propose an innovative model: SegKAN. First, we improve the conventional embedding module by adopting a novel convolutional network structure for image embedding, which smooths out image noise and prevents issues such as gradient explosion in subsequent stages. Next, we transform the spatial relationships between Patch blocks into temporal relationships to solve the problem of capturing positional relationships between Patch blocks in traditional Vision Transformer models. We conducted experiments on a Hepatic vessel dataset, and compared to the existing state-of-the-art model, the Dice score improved by 1.78%. These results demonstrate that the proposed new structure effectively enhances the segmentation performance of high-resolution extended objects. Code will be available at https://github.com/goblin327/SegKAN

Problem

• Image Fragmentation
• Noise Reduction
• Vessel Integrity

Innovation

• SegKAN
• Vascular Integrity
• Temporal Relation