Feature-reduction and semi-simulated data in functional connectivity-based cortical parcellation  

作　　者：Xiaoguang Tian Cirong Liu Tianzi Jiang Joshua Rizak Yuanye Ma Xintian Hu 

机构地区：[1]Kunming Institute of Zoology, Chinese Academy of Sciences [2]The University of Queensland, Queensland Brain Institute [3]LIAMA Center for Computational Medicine, National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences [4]Yunnan Key Lab of Primate Biomedical Research [5]State Key Laboratory of Brain and Cognitive Science, Institute of Biophysics, Chinese Academy of Sciences [6]Kunming Bio-International

出　　处：《Neuroscience Bulletin》2013年第3期333-347,共15页神经科学通报（英文版）

基　　金：supported by the National Basic Research Development Program (973 Program) of China (2012CBA01304, 2011CB707800);the National High Technology Research and Development Program (863 Program) of China (2012AA020701);the National Natural Science Foundation of China (31271167, 31271168, 81271495, 31070963, 31070965);the Strategic Priority Research Program of the Chinese Academy of Science, China (XDB02020500);the Development and Reform Project of Yunnan Province, China

摘　　要：Recently, restingstate functional magnetic resonance imaging has been used to parcellate the brain into functionally distinct regions based on the information available in functional connectivity maps. However, brain voxels are not independent units and adjacent voxels are always highly correlated, so functional connectivity maps contain redundant information, which not only impairs the computational efficiency during clustering, but also reduces the accuracy of clustering results. The aim of this study was to propose featurereduction approaches to reduce the redundancy and to develop semisimulated data with defined ground truth to evaluate these approaches. We proposed a featurereduction approach based on the Affinity Propagation Algorithm （APA） and compared it with the classic feature reduction approach based on Principal Component Analysis （PCA）. We tested the two approaches to the parcellation of both semisimulated and real seed regions using the Kmeans algorithm and designed two experiments to evaluate their noise resistance. We found that all functional connectivitymaps （with/without feature reduction） provided correct information for the parcellation of the semi simulated seed region and the computational efficiency was greatly improved by both feature reduction approaches. Meanwhile, the APAbased featurereduction approach outperformed the PCA based approach in noiseresistance. The results suggested that functional connectivity maps can provide correct information for cortical parcellation, and featurereduction does not significantly change the information. Considering the improvement in computational efficiency and the noiseresistance, featurereduction of functional connectivity maps before cortical parcellation is both feasible and necessary.Recently, restingstate functional magnetic resonance imaging has been used to parcellate the brain into functionally distinct regions based on the information available in functional connectivity maps. However, brain voxels are not independent units and adjacent voxels are always highly correlated, so functional connectivity maps contain redundant information, which not only impairs the computational efficiency during clustering, but also reduces the accuracy of clustering results. The aim of this study was to propose featurereduction approaches to reduce the redundancy and to develop semisimulated data with defined ground truth to evaluate these approaches. We proposed a featurereduction approach based on the Affinity Propagation Algorithm （APA） and compared it with the classic feature reduction approach based on Principal Component Analysis （PCA）. We tested the two approaches to the parcellation of both semisimulated and real seed regions using the Kmeans algorithm and designed two experiments to evaluate their noise resistance. We found that all functional connectivitymaps （with/without feature reduction） provided correct information for the parcellation of the semi simulated seed region and the computational efficiency was greatly improved by both feature reduction approaches. Meanwhile, the APAbased featurereduction approach outperformed the PCA based approach in noiseresistance. The results suggested that functional connectivity maps can provide correct information for cortical parcellation, and featurereduction does not significantly change the information. Considering the improvement in computational efficiency and the noiseresistance, featurereduction of functional connectivity maps before cortical parcellation is both feasible and necessary.

关 键 词：cortical parcellation resting-state fMRI functional connectivity feature reduction stimulateddata AP algorithm 

分 类 号：Q42[生物学—神经生物学] TP391.41[生物学—生理学]