About this Research Topic
The past decade has witnessed a great interest within the neuroscience community in modeling the function and structure of the brain and probing the dynamics between the two. DTI has been widely used to extract the structural network of the brain while the functional network has often been obtained via fMRI. Both linear and nonlinear models predicting the functional network from the structural one have been proposed. Such models potentially offer valuable tools in identifying abnormal regions of the brain. Graph theory has been exploited for the prediction of the structural network’s fiber atrophy spread due to brain diseases such as epilepsy and dementia, which can possibly help physicians determine and plan surgical as well as therapeutic interventions.
Graph signal processing, which is finding its way in computational neuroscience, analyzes time series emanating from the network nodes in the context of the underlying structural graph. Graph properties have been exploited in discerning characteristics of healthy as well as non-healthy brains, and although there have been breakthroughs in brain networks research, the landscape remains wide open. While there exists a body of works investigating the prediction of function from structure, by and large the literature has focused on fMRI as a functional modality with the implicit assumption of stationarity of the underlying time series. Alternate functional modalities such as EEG or MEG offer improved temporal resolution as compared to fMRI (at the cost of limited spatial resolution). Coupled with DTI or dMRI structural modalities and exploiting the time series’ non-stationarity, relating structure to improved temporal resolution function can shed alternative light on function/structure relationship for both healthy and non-healthy brains while incorporating the temporal delays resulting from high-frequency time series.
Another potential theme is the problem of estimating the structural graph from the functional one. Such an approach obviates the DTI step with its underestimated interhemispheric connections and can potentially lead to a more accurate estimate of the structural network.
The aims of this Research Topic are to advance the current understanding of the brain modeled as a mathematical network, develop new and advanced methods capturing the graph relationship between function and structure, propose new analysis methods of brain time series explicitly grounded on the underlying structural network, and ultimately apply the resulting knowledge to brain disease and further foster the understanding of the brain’s underlying configuration and dynamics.
We welcome submissions focusing on but not limited to the following topics:
- Application of tools such as machine learning, deep learning, and convolutional neural network to functional imaging data from fMRI to EEG to predict function from structure.
- Study of graph properties of structural or functional networks in healthy and non-healthy brains.
- Graph signal processing and harmonic analysis applied to brain networks.
- Novel methods and modalities for constructing functional and structural networks.
Graph signal processing, which is finding its way in computational neuroscience, analyzes time series emanating from the network nodes in the context of the underlying structural graph. Graph properties have been exploited in discerning characteristics of healthy as well as non-healthy brains, and although there have been breakthroughs in brain networks research, the landscape remains wide open. While there exists a body of works investigating the prediction of function from structure, by and large the literature has focused on fMRI as a functional modality with the implicit assumption of stationarity of the underlying time series. Alternate functional modalities such as EEG or MEG offer improved temporal resolution as compared to fMRI (at the cost of limited spatial resolution). Coupled with DTI or dMRI structural modalities and exploiting the time series’ non-stationarity, relating structure to improved temporal resolution function can shed alternative light on function/structure relationship for both healthy and non-healthy brains while incorporating the temporal delays resulting from high-frequency time series.
Another potential theme is the problem of estimating the structural graph from the functional one. Such an approach obviates the DTI step with its underestimated interhemispheric connections and can potentially lead to a more accurate estimate of the structural network.
The aims of this Research Topic are to advance the current understanding of the brain modeled as a mathematical network, develop new and advanced methods capturing the graph relationship between function and structure, propose new analysis methods of brain time series explicitly grounded on the underlying structural network, and ultimately apply the resulting knowledge to brain disease and further foster the understanding of the brain’s underlying configuration and dynamics.
We welcome submissions focusing on but not limited to the following topics:
- Application of tools such as machine learning, deep learning, and convolutional neural network to functional imaging data from fMRI to EEG to predict function from structure.
- Study of graph properties of structural or functional networks in healthy and non-healthy brains.
- Graph signal processing and harmonic analysis applied to brain networks.
- Novel methods and modalities for constructing functional and structural networks.
Keywords: Functional network, Structural network, Brain graph, Graph theory, Brain disease
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
