Dissociation Between Task Induced Variability Modulation And Evoked Activation Revealed By BOLD-fMRI

Recent blood oxygen level dependent functional magnetic resonance imaging(BOLD-fMRI)studies have found that the spatial patterns of task induced variability modulation(TIVM)were different from traditional task evoked activation(TEA),indicating that TIVM may provide a new perspective on human brain functional research.However,due to the methodological limitations,it was not clear whether the distinction indeed reflect neuronal response differences or not.Such question greatly limited the application of BOLD variability in brain imaging research.Therefore,the main purpose of current paper was to demonstrate that TIVM and TEA indeed reflect two distinct types of brain response through systematic studies and comparison.Specifically,the paper focused on the following two aspects:Study 1: the effects of data processing parameters to TIVM were evaluated,including global mean normalization(GMN)and band-pass filtering(BPF).Moreover,such analyses were repeated among three independent datasets to improve credibility.The results showed that the reproducibility of TIVM could be improved by GMN.TIVM was almost not affected by BPF.Study 2: the comparison between TIVM and TEA was performed both in space and time.Based on the methodological assessment in study 1,this study further compared TIVM and TEA from both in space and time.First,we used BOLD-fMRI and EO/EC paradigm to compare the spatial distinction between TIVM and TEA under same subjects and same experimental conditions.We found the increased TIVM in extrastriate cortex(ESC)and the decreased TIVM in primary sensorimotor cortex(PSMC),primary auditory cortex(PAC),supplementary motor area(SMA),and a few voxels in the thalamus in EO than EC whereas the TEA exhibited activation in primary visual cortex(PVC)and thalamus in EO compared to EC.By visual inspection,there was almost no spatial overlap between TIVM and TEA.Second,the temporal dynamics of TIVM were characterized by using expanding window(EW)technique.The results showed that TIVM exhibited a rather slow dynamical process,unlike the traditional TEA which could be measured in tens of seconds.In conclusion,GMN is beneficial to improve the sensitivity and reproducibility of TIVM.TIVM and TEA were dissociable both in space and time.Thus,we confirmed that TIVM and TEA indeed reflect two different neuronal response modes.Our results provides a basis for the application of TIVM in brain imaging studies.