Meeting Banner
Abstract #2143

Covariate Effects in Verbal Working Memory FMRI Using Schizophrenia and Control Participants

Kayako Matsuo1, S.H. Annabel Chen2, Chih-Min Liu3, Chen-Chung Liu3, Hai-Go Hwu3, Wen-Yih I. Tseng1

1National Taiwan University College of Medicine, Taipei, Taiwan; 2Division of Psychology, Nanyang Technological University, Singapore; 3Department of Psychiatry, National Taiwan University Hospital, Taipei, Taiwan

We investigated influences of various covariates on a verbal working memory (VWM) fMRI employing 18 each of schizophrenia and control participants. Six covariates of age, gender, response time (RT), accuracy rate, handedness score, and education year had all no significant differences between the two subject groups. Covariates by the behavioral data (RT and accuracy) specifically showed positive correlations with the DMN activation, which reflect the self-regulating mechanism during the low load condition. The inclusion of covariates into the design matrix was effective to obtain the “true” contrast effects even when the comparison groups were “matched” beforehand.

Abstract PDF Poster PDF

Keywords

accuracy activated activation addition affect almost alternation analyses array behavioral better biological blue button cerebellar challenges channel circuit cognition cognitive coil college computed conjunction conjunctions considered consonant cont contrast contrasts control controlled controls correlations cortex covariate covariates dealing decided default delay design designs despite diagnosed division duration dysfunction education effectiveness either employed employing ensure epochs examine experimental female frontal functional gender greatest hand handedness healthy honey identical illness included inclusion interleaved internal investigated kinds lateral letter letters load location lowercase maps matched matrix measures medial memory middle might mode national network neural nuisance options orthogonal overall overlaps paradigm parietal participants patients peak people positive preprocessing pressing previously price prime probe procedures processes psychiatry psychology quasi random reaction reflect relationship response scanner schizophrenia score section self significantly signs sixteen software specifically studied subject subjects suggests symptom systematically task thick tools transformation trio uncorrected uppercase usually variables variance various verbal year years yielded