Identity-mapping cascaded network for fMRI registration
Identity-mapping cascaded network for fMRI registration
复制标题
用于功能磁共振成像注册的身份映射级联网络
DOI：
10.1088/1361-6560/ac34b1
复制
发表时间：
2021-11-21
影响因子：
3.5
通讯作者：
Feng，Qianjin
中科院分区：
文献类型：
--
作者：
Zhu，Qiao Yun;Bai，HanHua;Feng，Qianjin
基于功能磁共振成像(FMRI)的神经科学研究依赖于功能区准确的受试者间图像配准。功能磁共振成像的被试间对齐可以提高群体分析的统计能力。最近的研究表明，基于深度学习的配准方法可以用于配准。在我们的工作中，我们提出了一种30个身份映射级联网络(30-IMCNet)用于RS-fMRI配准。它是一个级联网络，可以使运动图像逐渐扭曲，最终与固定图像对齐。在每个IMCNet的输入端增加一个带有身份映射路径的组合单元，用于指导网络训练。我们在RS-fMRI数据集(1000 Function Connectomes Project数据集)和与任务相关的fMRI数据集(Eyes Open Eyes Closed fMRI数据集)上实现了30-IMCNet。为了对我们的方法进行评估，我们在测试数据集中进行了组级分析。对于RS-fMRI，采用组水平t图的峰值t值、聚类水平评估和受试者间功能网络相关性等标准来评价注册质量。任务相关功能磁共振成像采用Alff配对t图的峰值t值和ReHo配对t图的峰值t值。与传统算法FSL、SPM和深度学习算法Kim等人相比，我们的方法在t-map的峰值t值上分别提高了48.90%、30.73%、36.38%和16.73%。我们提出的方法可以获得更好的功能注册性能，从而在功能一致性方面获得显着的改善。
Neuroscience researches based on functional magnetic resonance imaging (fMRI) rely on accurate inter-subject image registration of functional regions. The intersubject alignment of fMRI can improve the statistical power of group analyses. Recent studies have shown the deep learning-based registration methods can be used for registration. In our work, we proposed a 30-Identity-Mapping Cascaded network (30-IMCNet) for rs-fMRI registration. It is a cascaded network that can warp the moving image progressively and finally align to the fixed image. A Combination unit with an identity-mapping path is added to the inputs of each IMCNet to guide the network training. We implemented 30-IMCNet on an rs-fMRI dataset (1000 Functional Connectomes Project dataset) and a task-related fMRI dataset (Eyes Open Eyes Closed fMRI dataset). To evaluate our method, a group-level analysis was implemented in the testing dataset. For rs-fMRI, the criterions such as peak t-value of group-level t-maps, cluster-level evaluation, and intersubject functional network correlation were used to evaluate the quality of the registrations. For task-related fMRI, peak t-value in ALFF paired-t map and peak t-value in ReHo paired-t maps were used. Compared with traditional algorithm FSL, SPM, and deep learning algorithm Kim et al, Zhao et al our method has improvements of 48.90%, 30.73%, 36.38%, and 16.73% in the peak t value of t-maps. Our proposed method can achieve superior functional registration performance and thus gain a significant improvement in functional consistency.