The cluster-based test model for fMRI data analysis based on Neural Network

摘要：

Objective Accurately determining the statistical significance of changes in brain images is one of the most critical procedures of functional neuroimaging.Because of their increased sensitivity to spatially extended signals, cluster-size tests(CST) based on Random field theory (RFT) have been widely adopted in fMRI data analysis to detect brain activation.However, most existing CST approaches depend on the duster forming threshold and spatial smoothness level Thus, the performance of CST based on RFT is not stable.That”s because RFT relies on large assumptions and many parameters are estimated empirically ”1”.Neural network can develop an accurate model based on lots of learning information.Thus, we want to design a stable and powerful cluster-size test based on neural network under various smoothness levels and duster-forming threshold in this research.

作者: Huanjie Li Shen Lin Fengyu Cong

作者单位: Department of Biomedical Engineering, Dalian University of Technology, Liaoning, China, 116024

会议类型: 国内会议

会议名称: 第三届脑成像与脑电研究与应用研讨会

会议地点: 大连

会议语种:英文

页码: 34-34

在线出版日期: 2016-07-01（万方平台首次上网日期，不代表论文的发表时间）