Transient BOLD responses at block transitions
Introduction
Block design fMRI involves administering blocks of contrasting experimental conditions, for example task vs. rest, over epochs typically 25–40 s in duration. Conventionally, block design analyses evaluate signal modulation attributable to condition contrast assuming that the neuronal response is constant across each block. These techniques effectively collapse across time, thus discarding information in the temporal response profile. This analysis strategy dates back to the earliest functional neuroimaging studies based on positron emission tomography (PET) in which response modulations necessarily had to be measured over a sustained task period, usually about 1 min (Raichle et al., 1983). The greatly improved temporal resolution afforded by fMRI makes it possible to detect transient responses at block onset and offset in addition to sustained responses. Just as examination of the sustained response has provided insight into cognitive operations occurring throughout a task block, examination of transient responses could provide important information about functional processes occurring specifically at the beginning or end of a task block (block transitions). This idea is reinforced by recent work indicating that transient BOLD responses at block transitions can be specifically modified by clinical state (e.g., diagnosis of schizophrenia) (Fox et al., 2005). Such data highlight the need for an increased understanding of transient BOLD responses at block transitions.
Transient phenomena at the onset of the task-related BOLD response have been noted for many years but, in the preponderance of extant studies, have been attributed to transient mismatch between neuronal metabolic demand (assumed to be constant within task blocks) and various components of the vascular response (Buxton et al., 1998, Obata et al., 2004). In other words, BOLD transients at block transitions have generally not been viewed as reflecting transient neuronal activity. A notable exception is the study by Konishi et al. (2001), which found that intentional encoding task block transients exhibited a spatial distribution distinct from that of the sustained response, suggesting that transient and sustained components of the BOLD signal might reflect distinct functional processes. Their study, however, was limited to a single task paradigm and made no distinction between the transient response at task onset vs. offset (Konishi et al., 2001). We here present a meta-analysis of BOLD transient responses distinguishing between task onset and offset using four very different task conditions: 2-back working memory, visual scene analysis, overt word reading, and simple visual fixation. We address four major questions:
- 1.
Can transient vs. sustained components in block design fMRI experiments be reliably distinguished and localized?
- 2.
Which regions exhibit the largest transient responses on average? Are there regions that always show transient responses at task state transitions?
- 3.
To what extent do transient BOLD responses depend on task? Are these responses different at block onset and offset within the same task?
- 4.
How do these empirical observations constrain our interpretation of the functional role of transient activity at block transitions?
It should be noted that the current study is an analysis of transient BOLD responses at the onset and offset of a task block. This is different from recently published analyses which examine state and item effects (sometimes referred to as BOLD transients) within a task block (Velanova et al., 2003, Visscher et al., 2003).
Section snippets
Methods
The current analysis is based on data acquired in four previous block design fMRI studies not originally focused on block transition transients. These data were selected to include a broad range of cognitive activities. A summary listing of task and control conditions, subject number, stimuli, scanning parameters, and original sources of these data is given in Table 1.
Separation of onset transient, sustained, and offset transient responses
For each of the four tasks, voxel-based maps of significant onset transient, sustained, and offset transient responses were created. Results for one of the tasks, the working memory task, are shown in Fig. 2. To illustrate face validity of the GLM based procedure for separation of transient and sustained components, single-slice square regions of interest (not to be confused with the ROIs listed in Table 2) are shown overlaid on the distributions. The time courses for a single epoch (40 s task
Separation of transient and sustained components
We have presented a GLM-based approach for analysis of transient and sustained responses in block design studies. In addition to the sustained BOLD increases measured in most conventional block-design analysis strategies, the current technique can be used to localize significant transient responses at block transitions. Importantly, the current technique obtains an estimate of the transient without assuming the shape of the response. We demonstrated face validity of our technique by showing
Acknowledgments
We thank Mark McAvoy for assistance with data analysis and statistics. This work was supported by grant NS06833.
References (59)
The Madrid card sorting task (MCST): a task switching paradigm to study executive attention with event-related potentials
Brain Res. Protoc.
(2003)- et al.
A parametric study of prefrontal cortex involvement in human working memory
Neuroimage
(1997) - et al.
Neural mechanisms of transient and sustained cognitive control during task switching
Neuron
(2003) - et al.
Prefrontal cortex activation in task switching: an event related fMRI study
Cogn. Brain Res.
(2000) - et al.
The BOLD onset transient: identification of novel functional differences in schizophrenia
NeuroImage
(2005) - et al.
Analysis of fMRI time-series revisited
Neuroimage
(1995) - et al.
Stimulus-dependent BOLD and perfusion dynamics in human V1
Neuroimage
(1999) - et al.
Generalisability, random effects and population inference
NeuroImage
(1998) - et al.
Modulation of task-related neural activity in task-switching: an fMRI study
Cogn. Brain Res.
(2000) - et al.
Transient activation during block transition
Neuroimage
(2001)