摘要
对一种多检验方法———错误发现率 (FDR)方法进行了计算机实现 ,并将FDR方法及其他几种假设检验方法应用在模拟数据和fMRI数据分析中进行比较 。
One of the important steps to analyze brain imaging data is to determine the activation region of brain by hypothesis test. Single test method and some classical multiple test methods can not give out a good solution to multiple test problem in analyzing brain imaging data. A new multiple test method, false discovery rate (FDR) method, can get best result by controlling the rate of falsely rejected hypothesis. FDR method and other test methods are used to analyze simulation and fMRI data. The results prove that FDR method is a more creditable multiple test method.
出处
《北京师范大学学报(自然科学版)》
CAS
CSCD
北大核心
2003年第3期422-426,F003,共6页
Journal of Beijing Normal University(Natural Science)
基金
国家自然科学基金资助项目 (6 0 0 72 0 0 3)