TY - GEN
T1 - A Multi-Modal Fusion Method for 3-Hinge Gyrus Based on Bidirectional Cross-Attention
AU - Jia, Chenjie
AU - Li, Xiao
AU - Sun, Qitai
AU - Zhang, Han
AU - He, Xiaowei
AU - Zhang, Tuo
AU - Ren, Yudan
N1 - Publisher Copyright:
© 2025 IEEE.
PY - 2025
Y1 - 2025
N2 - The 3-hinge Gyrus (3HG) is the junction of gyri from three directions during cortical folding, exhibiting unique and consistent structural and functional characteristics across individuals and species. Its number, location, and morphology can be used for quantitative analysis of cortical folding patterns. However, it is worth noting that most studies extract and quantify 3HG features on different modality respectively, which is limited in capturing the complementary information between different modalities. Thus, we propose a multimodal fusion method based on bidirectional cross-attention to integrate structural MRI and DTI data in this work. Results demonstrate that our method not only preserves the unique features of each modality, also is effective in capturing cross-modality features which are unavailable in single-modality analysis, revealing connectivity and potential functional relationships between brain regions. This multimodal fusion method may provide a further understanding on mechanisms of cortical folding patterns and brain network.
AB - The 3-hinge Gyrus (3HG) is the junction of gyri from three directions during cortical folding, exhibiting unique and consistent structural and functional characteristics across individuals and species. Its number, location, and morphology can be used for quantitative analysis of cortical folding patterns. However, it is worth noting that most studies extract and quantify 3HG features on different modality respectively, which is limited in capturing the complementary information between different modalities. Thus, we propose a multimodal fusion method based on bidirectional cross-attention to integrate structural MRI and DTI data in this work. Results demonstrate that our method not only preserves the unique features of each modality, also is effective in capturing cross-modality features which are unavailable in single-modality analysis, revealing connectivity and potential functional relationships between brain regions. This multimodal fusion method may provide a further understanding on mechanisms of cortical folding patterns and brain network.
KW - 3-hinge Gyrus
KW - Brain connectivity network
KW - Cortical folding patterns
KW - Multi-modal data fusion
UR - http://www.scopus.com/inward/record.url?scp=105005824887&partnerID=8YFLogxK
U2 - 10.1109/ISBI60581.2025.10980730
DO - 10.1109/ISBI60581.2025.10980730
M3 - 会议稿件
AN - SCOPUS:105005824887
T3 - Proceedings - International Symposium on Biomedical Imaging
BT - ISBI 2025 - 2025 IEEE 22nd International Symposium on Biomedical Imaging, Proceedings
PB - IEEE Computer Society
T2 - 22nd IEEE International Symposium on Biomedical Imaging, ISBI 2025
Y2 - 14 April 2025 through 17 April 2025
ER -