Abstract: ObjectiveTo explore fMRI data with independent component analysis (ICA) in order toinvestigate effects of transcu-taneous electrical acupoint sti mulation (TEAS) on brain function.MethodsThe experi ment was performed on a whole-body 1 .5 TGE Signa Excite MRI scanner with which the brain oxygenation level dependent (BOLD)/EPI i mages were acquired froma femaletraumatic braininjury patient . A block designed protocol was used . Both durations of rest and TEAS were 30 seconds . The dataprocessing was performed with GIFT,Statistical Parametric Mapping 5 (SPM5) and MRIcro .Results fromICAand SPMwere com-pared .Results Extended Infomax algorithmprovided by GIFT found thirteen independent components (ICs) ,each of which containeda spatial map and a corresponding ti me course . The spatial maps associated with task-related ICs resembled the activation maps from SPM5 but were not totally identical .In addition,the ti me courses of these ICs differed from the shape of canonical HRF model usedby SPM.Conclusion ICA is a good choice to investigate data and obtain prior knowledge before using model-based methods such as SPM.

Key words: independent component analysis (ICA), transcutaneous electrical acupoint stimulation (TEAS), functional magnetic resonance imaging (fMRI), statistical parametric mapping (SPM)