Complex spatio-temporal dynamics of fMRI BOLD: A study of motor learning |
| |
Authors: | Duff Eugene Xiong Jinhu Wang Binquan Cunnington Ross Fox Peter Egan Gary |
| |
Affiliation: | The Howard Florey Institute and the Centre for Neuroscience, The University of Melbourne, VIC 3010, Australia. eduff@pcomm.hfi.unimelb.edu.au |
| |
Abstract: | Many studies have investigated the temporal properties of BOLD signal responses to task performance in regions of interest, often noting significant departures from the conventionally modelled response shape, and significant variation between regions. However, these investigations are rarely extended across the whole brain nor incorporated into the routine analysis of fMRI studies. As a result, little is known about the range of response shapes generated in the brain by common paradigms. The present study finds such temporal dynamics can be complex. We made a detailed investigation of BOLD signal responses across the whole brain during a two minute motor-sequence task, and tracked changes due to learning. The multi-component OSORU (Onset, Sustained, Offset, Ramp, Undershoot) linear model, developed by Harms and Melcher (J.Neurophysiology, 2003), was extended to characterise responses. In many regions, signal transients persisted for over thirty seconds, with large signal spikes at onset often followed by a dip in signal below the final sustained level of activation. Training altered certain features of the response shape, suggesting that different features of the response may reflect different aspects of neuro-vascular dynamics. Unmodelled, this may give rise to inconsistent results across paradigms of varying task durations. Few of the observed effects have been thoroughly addressed in physiological models of the BOLD response. The complex, extended dynamics generated by this simple, often employed task, suggests characterisation and modelling of temporal aspects of BOLD responses needs to be carried out routinely, informing experimental design and analysis, and physiological modelling. |
| |
Keywords: | |
本文献已被 ScienceDirect PubMed 等数据库收录! |
|