Identification of Cerebral Networks by Classification of the Shape of BOLD Responses

• Published in: Journal of neurophysiology 2003-07, Vol.90 (1), p.360-371
• Main Authors: d'Avossa, Giovanni, Shulman, Gordon L, Corbetta, Maurizio
• Format: Article
• Language: English
• Subjects: Brain - physiology; Brain Mapping; Cerebrovascular Circulation - physiology; Cluster Analysis; Color Perception - physiology; Cues; Evoked Potentials, Visual - physiology; Humans; Image Processing, Computer-Assisted; Magnetic Resonance Imaging - methods; Motion Perception - physiology; Oxygen - blood; Photic Stimulation; Time Factors; Visual Perception - physiology
• ISSN: 0022-3077; 1522-1598
• DOI: 10.1152/jn.01040.2002

1 Department of Neurology and Neurological Surgery, 2 Alzheimer's Disease Research Center, 3 Department of Radiology, and 4 Department of Anatomy and Neurobiology, Washington University, St. Louis, Missouri 63110 Submitted 18 November 2002; accepted in final form 3 December 2002 Changes in regional blood oxygen level dependent (BOLD) signals in response to brief visual stimuli can exhibit a variety of time-courses. To demonstrate the anatomical distribution of BOLD response shapes during a match to sample task, a formal analysis of their time-courses is presented. An event-related design was used to estimate regional BOLD responses evoked by a cue word, which instructed the subject to attend to the motion or color of an upcoming target, and those evoked by a briefly presented moving target consisting of colored dots. Regional BOLD time-courses were adequately represented by the linear combination of three orthogonal waveforms. BOLD response shapes were then classified using a fuzzy clustering scheme. Three classes (sustained, phasic, and negative) best characterized cue responses. Four classes (sustained, sustained-phasic, phasic, and bi-phasic) best characterized target responses. In certain regions, the shape of the BOLD responses was modulated by the instruction to attend to the target's motion or color. A left frontal and a posterior parietal region showed sustained activity when motion was cued and transient activity when color was cued. A right thalamic and a left lateral occipital region showed sustained activity when color was cued and transient activity when motion was cued. Following the target several regions showed more sustained activity during motion than color trials. In summary, the effect of the task variable was focal following the cue and widespread following the target. We conclude that the temporal patterns of neural activity affected the shape of the BOLD signal. Address for reprint requests: G. d'Avossa, Alzheimer's Disease Research Center, Washington University School of Medicine, 4488 Forest Park Ave., Suite 130, St. Louis, MO 63108 (E-mail: davossag{at}neuro.wustl.edu ).