Assessing linear time‐invariance in human primary somatosensory cortex with BOLD fMRI using vibrotactile stimuli

Abstract

The assumption of linear time-invariance (LTI) in the human primary somatosensory cortex (SI) is assessed for fMRI signals generated by variable-duration vibrotactile stimuli. Predictions based on time-shifted summation (TSS) of responses to 2 s stimuli overestimate observed BOLD signal amplitudes in response to longer-duration stimuli, in agreement with previous findings in other primary sensory cortices. To interpret these results, we undertook an alternative approach for LTI assessment by characterizing BOLD signals using two biophysical models. The first model assumes that the input stimulus envelope is proportional to neural activity. The second assumes that neural activity exhibits both transient and steady-state components, consistent with extensive electrophysiological data, and fits the experimental data better. Although nonlinearity remains evident for short stimulus durations, the latter model shows that the TSS procedure to assess LTI overestimates the BOLD signal because the temporal characteristics of neural activity have not been considered adequately.

Transient and sustained components of the sensorimotor BOLD response in fMRI

“…1-5 are also consistent with the notion, previously stated by others [5,8,10,12,19,20,22], that nonlinearity in BOLD fMRI studies associated with short sensory stimulus durations can be accounted for by a superposition of linear processes represented here by the transient and sustained components of the neural input function. It is possible that a nonlinear model, potentially including a nonlinear hemodynamic response, for example, would also be able to explain the data.…”

“…This is in agreement with the findings of Nangini et al [20] that the fMRI response to 0.2-s, 0.5-s and 1-s stimuli is dominated by transient responses. An earlier study by Nangini et al [19] investigated stimuli up to 20 s in duration and found that a model with an onset transient and a boxcar component fit the observed BOLD responses better than a pure boxcar model. While a pure transient model was not investigated in this prior study, the sustained, boxcar component was easily visible at such a long stimulus duration.…”

“…Of particular interest is S1c, the primary cortical area responding to a contralateral somatosensory stimulus [19,20,29], which has been shown to behave similarly to other primary sensory areas with respect to the linearity of the BOLD response as a function of stimulus duration [5]. In the context of our models, the boxcar component is the conventional function used in GLM analysis of fMRI data and for assessments of linear time invariance.…”

“…However, significant deviations from LTI properties have been reported for BOLD responses of the visual [5,6,[8][9][10][11][12][13][14][15][16], motor [5,9,10,17], auditory [5,17,18] and somatosensory systems [19,20] for stimulus durations less than approximately 3 s. Moreover, such deviations have also been found for cognitive tasks [21]. Recognizing that the assumption of constant neural activity for a sustained stimulus is an oversimplification, a number of authors have suggested that neural transients or neural adaptation could explain, at least partially, such deviations from LTI properties [5,7,8,10,12,[19][20][21][22]. These reports primarily focus on transient stimulus onset responses.…”

“…Previously, Nangini et al investigated the primary somatosensory responses to passively experienced vibrotactile stimulation with fMRI and MEG [19,20] and suggested to consider transient neural components. In these experiments, the fMRI signal showed substantial intersubject variability that was thought to occur at least partly due to fluctuations in attention.…”