k-Uniform Hypergraph Reasoning Network for Focal Liver Lesions Classification in Multimodal Magnetic Resonance Imaging

Shaocong Mo, Ming Cai, Lanfen Lin, Ruofeng Tong, Fang Wang, Qingqing Chen, Hongjie Hu, Yinhao Li, Yen-Wei Chen

Abstract: Accurately identifying of focal liver lesions (FLLs) is crucial for liver diagnostics. In clinical, multimodal magnetic resonance imaging (MRI) offers comprehensive visual patterns vital for differentiating various liver tumors. In this paper, we introduce a uniform hypergraph based reasoning network for the classification of FLLs in multimodal MRI. Our model employs a shared-weight encoder for feature extraction and then constructs a uniform hypergraph with hyper-parameter k to capture high-order relationship in a group-wise manner. Iterative message passing within the hypergraph refines node features by integrating complementary information from the connected nodes of each hyperedge. The effectiveness of our proposed method was evaluated on the LLD-MMRI2023 dataset and demonstrated better performance over conventional approaches.