Fusion of FMRI-sMRI-EEG by Ensemble Feature Selection Improves Classification of Schizophrenia

	Fusion of FMRI-sMRI-EEG by Ensemble Feature Selection Improves Classification of Schizophrenia
	Sui Jing(隋婧); Hao He; Yuhui Du; Qingbao Yu; Jiayu Chen; Eduardo Castro; David A Bridwell; Godfrey D Pearlson; Vince D Calhoun
	2014
会议名称	2014 The 20th annual confernece of Organization for Human Brain Mapping(OHBM 2014)
会议日期	2014/6/8-12
会议地点	Hamburg, Germany
摘要	Nonuniformity correction (NUC) is a critical task for achieving higher performances in modern infrared imaging systems. For cases where radiometry is not required, we proposed an extension to a recently reported scene-based NUC technique RASBA, the scene-based algorithm using perimeter diaphragm strips (SBA-PDS)*. This method initially guarantees all detectors along FPA perimeter have an uniform, but unknown bias through one-point calibration, which is dependent on the reciprocating movement of diaphragm strips. Then the SBA-PDS proceeds bias estimation recursively based on a special algebraic algorithm and can effectively "transport" the calibration of the perimeter detectors to those interior uncorrected ones. This approach provides the advantages of operating NUC with an almost unobstructive field of view, no need for cost of blackbody sources, and achieving acceptable results during the time of hundreds of frames, which is usually as long as thousands of frames by statistical algorithm. The technique was applied to real infrared data obtained from two kinds of uncooled infrared cameras and the experimental results appeared promising.
关键词	Fmri-smri-eeg Feature Selection Improves Classification Of Schizophrenia
文献类型	会议论文
条目标识符	http://ir.ia.ac.cn/handle/173211/20786
专题	脑图谱与类脑智能实验室_脑网络组研究
作者单位	Institute of Automation Chinese Academy of Sciences
第一作者单位	中国科学院自动化研究所
推荐引用方式 GB/T 7714	Sui Jing,Hao He,Yuhui Du,et al. Fusion of FMRI-sMRI-EEG by Ensemble Feature Selection Improves Classification of Schizophrenia[C],2014.