[1]索国栋,周月娥,金丰庆,等.基于模态可靠性与可形变跨尺度共注意力的多模态脑肿瘤MRI分割[J].中国医学物理学杂志,2026,43(7):930-938.[doi:DOI:10.3969/j.issn.1005-202X.2026.07.014]
　SUO Guodong,ZHOU Yuee,et al.Multi-modal brain tumor MRI segmentation based on modality reliability and deformable cross-scale co-attention[J].Chinese Journal of Medical Physics,2026,43(7):930-938.[doi:DOI:10.3969/j.issn.1005-202X.2026.07.014]
点击复制

基于模态可靠性与可形变跨尺度共注意力的多模态脑肿瘤MRI分割()

《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]

卷:
43卷
期数:
2026年第7期
页码:
930-938
栏目:
医学影像物理
出版日期:
2026-07-22

文章信息/Info

Title:
Multi-modal brain tumor MRI segmentation based on modality reliability and deformable cross-scale co-attention
文章编号:
1005-202X（2026）07-0930-09
作者:
索国栋1; 2; 周月娥1; 2; 金丰庆1; 2; 杨建兰2
1.甘肃中医药大学医学信息工程学院， 甘肃 兰州 730000； 2.泉州市正骨医院， 福建 泉州 362000
Author(s):
SUO Guodong1;  2;  ZHOU Yuee1;  2;  JIN Fengqing1;  2;  YANG Jianlan2
1. School of Medical Information Engineering, Gansu University of Chinese Medicine, Lanzhou 730000, China 2. Quanzhou Orthopedic-Traumatological Hospital, Quanzhou 362000, China
关键词:
脑肿瘤; 图像分割; 模态可靠性感知融合; 跨尺度可形变共注意力; 不确定性引导深监督
Keywords:
Keywords: brain tumor image segmentation modality reliability-aware fusion deformable cross-scale co-attention uncertainty-guided deep supervision
分类号:
R318
DOI:
DOI:10.3969/j.issn.1005-202X.2026.07.014
文献标志码:
A
摘要:
目的：在Medical Imaging Segmentation Toolkit集成的U-Net骨干基础上提出一种基于模态可靠性与可形变跨尺度共注意力的三维（3D）多模态脑肿瘤分割网络。方法：设计3项轻量级改进：在输入端引入模态可靠性感知融合模块，自适应抑制低质量模态噪声；在解码阶段构建可形变跨尺度共注意力模块，增强肿瘤边界及形变区域的表达；在深监督分支中加入不确定性引导深监督模块，提升对小病灶和模糊区域的优化能力。结果：在BraTS2021数据集上进行验证，与多种3D分割基线模型进行比较。结果发现本文模型在完整肿瘤、肿瘤核心和增强肿瘤区域均取得性能领先的效果：完整肿瘤区的Dice相似性系数（DSC）为0.921，肿瘤核心区的DSC为0.895，增强肿瘤区的DSC为0.891，平均95% Hausdorff距离为4.122 mm，优于其他模型。结论：本研究提出的3D多模态脑肿瘤分割网络在保持网络结构轻量化的前提下，有效提升多模态信息融合质量、跨尺度特征对齐能力及深监督优化效率，为脑肿瘤MRI自动分割提供一种具有鲁棒性和精细度的方案。
Abstract:
Abstract: Objective To propose a three-dimensional (3D) multi-modal brain tumor segmentation network built upon the U-Net backbone implemented in the Medical Imaging Segmentation Toolkit, which integrates modality reliability and deformable cross-scale co-attention. Methods Three lightweight improvements were designed for the network. Specifically, a modality reliability-aware fusion module was embedded at the input stage to adaptively suppress low-quality modality noise. A deformable cross-scale co-attention module was introduced within the decoder to enhance the representation of tumor boundaries and structurally deformed regions. An uncertainty-guided deep supervision module was used in the deep supervision branch to improve the optimization performance for small lesions and ambiguous regions. Results The proposed method was validated on the BraTS2021 dataset and compared with several 3D segmentation baseline models. Results showed that the proposed model achieved optimal performance across whole tumor, tumor core, and enhancing tumor subregions, with Dice similarity coefficients of 0.921, 0.895, and 0.891, respectively, and a mean 95% Hausdorff distance reaching 4.122 mm, which outperformed the other models. Conclusion The proposed 3D multi-modal brain tumor segmentation network effectively enhances multi-modal feature fusion, cross-scale feature alignment, and deep supervision optimization, while maintaining a lightweight architecture. This study provides a robust and fine-grained solution for automated brain tumor segmentation from MRI images.

相似文献/References:

[1]陈海斌,甄 鑫,周凌宏.基于先验的随机游走算法在医学图像分割中的应用[J].中国医学物理学杂志,2015,32(02):174.[doi:10.3969/j.issn.1005-202X.2015.02.005]
[2]王晓飞,聂生东,王远军,等.基于MRI的脑肿瘤计算机辅助检测技术研究进展[J].中国医学物理学杂志,2014,31(01):4635.[doi:10.3969/j.issn.1005-202X.2014.01.010]
[3]许红玉,蔡坦坦,叶良凯,等.分水岭算法在CT图像分割中的应用[J].中国医学物理学杂志,2014,31(06):5272.[doi:10.3969/j.issn.1005-202X.2014.06.012]
[4]万俊,聂生东,王远军,等.基于MRI的脑肿瘤分割技术研究进展[J].中国医学物理学杂志,2013,30(04):4266.[doi:10.3969/j.issn.1005-202X.2013.04.013]
[5]周露,张书旭,余辉,等.PET-CT图像配准的预处理研究[J].中国医学物理学杂志,2013,30(05):4392.[doi:10.3969/j.issn.1005-202X.2013.05.012]
[6]谷珊珊,张怀文,王运来.基于水平集稳健特征统计算法的脑肿瘤自动分割研究[J].中国医学物理学杂志,2016,33(1):63.[doi:10.3969/j.issn.1005-202X.2016.01.014]
　[J].Chinese Journal of Medical Physics,2016,33(7):63.[doi:10.3969/j.issn.1005-202X.2016.01.014]
[7]刘国才,胡泽田,朱苏雨,等.头颈部肿瘤PET与MRI融合放疗靶区自适应区域生长勾画[J].中国医学物理学杂志,2016,33(3):222.[doi:10.3969/j.issn.1005-202X.2016.03.002]
　[J].Chinese Journal of Medical Physics,2016,33(7):222.[doi:10.3969/j.issn.1005-202X.2016.03.002]
[8]贾高杰,邱崧,蔡茗名,等.三维点云重构和体显示在医学辅助诊断中的应用[J].中国医学物理学杂志,2016,33(6):593.[doi:10.3969/j.issn.1005-202X.2016.06.012]
　[J].Chinese Journal of Medical Physics,2016,33(7):593.[doi:10.3969/j.issn.1005-202X.2016.06.012]
[9]徐艳红,张晋建,黎静,等.影像增强剂对脑瘤螺旋断层放疗剂量计算的影响[J].中国医学物理学杂志,2016,33(8):757.[doi:10.3969/j.issn.1005-202X.2016.08.001]
　[J].Chinese Journal of Medical Physics,2016,33(7):757.[doi:10.3969/j.issn.1005-202X.2016.08.001]
[10]张泽凡,张东.基于交叉熵和GVF-Snake的子宫肌瘤高强度聚焦超声图像自动分割算法[J].中国医学物理学杂志,2016,33(8):776.[doi:10.3969/j.issn.1005-202X.2016.08.005]
　[J].Chinese Journal of Medical Physics,2016,33(7):776.[doi:10.3969/j.issn.1005-202X.2016.08.005]
[11]赖胜圣,刘虔铖,张刚平.基于模糊C均值自动随机游走算法在脑肿瘤分割中的应用[J].中国医学物理学杂志,2015,32(05):707.[doi:doi:10.3969/j.issn.1005-202X.2015.05.021]
[12]刘遵雄,陈子涵,蔡体健,等.基于多尺度细节增强的脑瘤图像分割研究[J].中国医学物理学杂志,2024,41(7):828.[doi:DOI:10.3969/j.issn.1005-202X.2024.07.007]
　LIU Zunxiong,CHEN Zihan,CAI Tijian,et al.Brain tumor image segmentation based on multi-scale detail enhancement[J].Chinese Journal of Medical Physics,2024,41(7):828.[doi:DOI:10.3969/j.issn.1005-202X.2024.07.007]

备注/Memo

备注/Memo:
【收稿日期】2026-03-21 【基金项目】泉州市科技计划项目（2026QZNY048） 【作者简介】索国栋，硕士研究生，研究方向：医学图像分割与深度学习，E-mail: liupizhi666@163.com 【通信作者】杨建兰，副教授，研究方向：医学影像识别与应用、卫生信息管理系统开发与研究、卫生信息数据挖掘，E-mail: fjyjl@gszy.edu.cn
更新日期/Last Update: 2026-07-22