Functional neuroimaging, and in particular fMRI, has shown much promise in uncovering brain-behavior relationships, with a recent focus on functional connectivity. With regards to the latter, most functional imaging studies have tended to aggregate information across an entire four-dimensional scan into a one- or two-dimensional measure, such as an adjacency matrix describing functional connectivity (FC) of the brain. This approach, however, discards much of the rich temporal information present in the data and ignores the dynamic aspect of FC (termed dynamic FC – dFC) over different time scales, which may yield important, cognitively relevant markers.
In this context, new methods to estimate dFC have emerged, aiming to accurately track FC variation patterns from an individual’s scan as much as possible. Specifically, it has been shown that FC topology varies on timescales much shorter than a typical scan length, and that these changes may reflect inter-subject trait differences. It has also been suggested that dFC patterns, captured in as little as tens of seconds, reflect changes in vigilance and cognitive engagement, human error in a visual task, and cognitive decline in elderly populations. Importantly, differences in dFC patterns between healthy and pathological (e.g. schizophrenic) participants have been also reported.
While dFC studies have shown significant potential, dFC estimation is characterized by lower SNR, poorer reliability, and higher susceptibility to noise processes overall. This is particularly the case for fMRI data, when compared to modalities such as EEG, due to the relatively low number of observations. In addition, disentangling the purely neural component of fMRI-based dFC is challenging, as it is well known that the fMRI signal is affected by many additional physiological processes in a regionally dependent manner. Related to this, validation of dFC patterns in resting state fMRI can be difficult, as we often lack a ground truth for determining the factors that spur changes in FC.
The validity of resting-state dFC metrics, and their link with underlying neural processes, can be investigated in several ways. One way is through the use of tasks or stimuli, as more reproducible, synchronous changes in dFC can be driven by a common task across participants. Furthermore, task paradigms can drive robust differences in brain states, which can be leveraged to validate the performance of dFC methods. Another way is to follow a multimodal approach, whereby dFC patterns can be more reliably estimated and interpreted by integrating fMRI with complementary modalities, such as EEG, Calcium Imaging, noninvasive brain stimulation (e.g. TMS), and/or pupillometry and other physiological recordings.
The present research topic will accept papers which attempt to improve the assessment, understanding, and validation of dFC using approaches such as task-based or multimodal functional neuroimaging, in human or animal studies. Studies that aim to leverage dFC to investigate and shed light on brain-behavior relationships are of particular interest. We will focus on the following:
- Brain state manipulation: Task based functional imaging or (noninvasive) brain stimulation can drive reliable changes in brain connectivity, and provide a reference for interpreting dFC metrics
- Statistical modelling: Validation of experimental findings using null models, test-retest reliability using longitudinal/repeat measurements, or out of sample replication
- Multimodal studies: Functional neuroimaging using multiple modalities, incorporation of physiological recordings in dFC assessment, or a combination thereof
Functional neuroimaging, and in particular fMRI, has shown much promise in uncovering brain-behavior relationships, with a recent focus on functional connectivity. With regards to the latter, most functional imaging studies have tended to aggregate information across an entire four-dimensional scan into a one- or two-dimensional measure, such as an adjacency matrix describing functional connectivity (FC) of the brain. This approach, however, discards much of the rich temporal information present in the data and ignores the dynamic aspect of FC (termed dynamic FC – dFC) over different time scales, which may yield important, cognitively relevant markers.
In this context, new methods to estimate dFC have emerged, aiming to accurately track FC variation patterns from an individual’s scan as much as possible. Specifically, it has been shown that FC topology varies on timescales much shorter than a typical scan length, and that these changes may reflect inter-subject trait differences. It has also been suggested that dFC patterns, captured in as little as tens of seconds, reflect changes in vigilance and cognitive engagement, human error in a visual task, and cognitive decline in elderly populations. Importantly, differences in dFC patterns between healthy and pathological (e.g. schizophrenic) participants have been also reported.
While dFC studies have shown significant potential, dFC estimation is characterized by lower SNR, poorer reliability, and higher susceptibility to noise processes overall. This is particularly the case for fMRI data, when compared to modalities such as EEG, due to the relatively low number of observations. In addition, disentangling the purely neural component of fMRI-based dFC is challenging, as it is well known that the fMRI signal is affected by many additional physiological processes in a regionally dependent manner. Related to this, validation of dFC patterns in resting state fMRI can be difficult, as we often lack a ground truth for determining the factors that spur changes in FC.
The validity of resting-state dFC metrics, and their link with underlying neural processes, can be investigated in several ways. One way is through the use of tasks or stimuli, as more reproducible, synchronous changes in dFC can be driven by a common task across participants. Furthermore, task paradigms can drive robust differences in brain states, which can be leveraged to validate the performance of dFC methods. Another way is to follow a multimodal approach, whereby dFC patterns can be more reliably estimated and interpreted by integrating fMRI with complementary modalities, such as EEG, Calcium Imaging, noninvasive brain stimulation (e.g. TMS), and/or pupillometry and other physiological recordings.
The present research topic will accept papers which attempt to improve the assessment, understanding, and validation of dFC using approaches such as task-based or multimodal functional neuroimaging, in human or animal studies. Studies that aim to leverage dFC to investigate and shed light on brain-behavior relationships are of particular interest. We will focus on the following:
- Brain state manipulation: Task based functional imaging or (noninvasive) brain stimulation can drive reliable changes in brain connectivity, and provide a reference for interpreting dFC metrics
- Statistical modelling: Validation of experimental findings using null models, test-retest reliability using longitudinal/repeat measurements, or out of sample replication
- Multimodal studies: Functional neuroimaging using multiple modalities, incorporation of physiological recordings in dFC assessment, or a combination thereof
