Cognitive neuroscience has traditionally focused on the design of simple tasks to isolate a particular cognitive domain for investigation. While this form of research has yielded fundamental knowledge on the functional specialization in the brain, its ecological validity remains unclear. Accordingly, a recent development in neuroimaging is to design ecologically valid naturalistic paradigms that approximate real-life scenarios, using naturalistic and dynamic stimuli such as films and music. Comparing to traditional task-based literature, studying neural processing and network interactions during natural, dynamic conditions is imperative to the understanding of human brain function, further offering a powerful tool for clinical research on brain disorders, for their high compliance and low demand.
While naturalistic paradigms are increasingly used to map brain functions in healthy populations or altered brain functions in clinical populations, one challenge is the lack of effective computational models to characterize neural correlates of naturalistic stimuli. To date, various methods have been successfully applied to characterize different aspects of brain activities under naturalistic paradigms, such as General Linear Model (GLM), and Independent Component Analysis (ICA), and Inter-subject Correlation (ISC), etc. However, there are still many challenges in properly modeling the spatial-temporal characteristics and hidden features from these high-dimensional neuroimaging data with multimodal, dynamic naturalistic stimuli.
Recently, data mining and modeling methods have developed in machine learning and artificial intelligence, gaining increasing attention in the computational neuroscience field. These innovative methods can be nicely positioned to exploit big neuroimaging data sets by tackling the complex challenges presented by multimodal contents of the stimuli.
The aim of this Research Topic is to provide an interdisciplinary platform for researchers to investigate the computational methods for modeling neuroimaging data (fMRI, EEG, ECoG, MEG, etc.) that employ the rich, multimodal dynamic stimuli, such as film clips, TV advertisements, news items, and spoken narratives, or that embody relatively unconstrained interactions with other agents, gaming environments, or virtual realities.
We welcome original research articles and review articles that focus on one of the three following aspects, or all:
1. Methodological advances in the analysis of neuroimaging data using naturalistic paradigm (such as fMRI, EEG, MEG, etc) that allow the identification of dynamic functional brain states or functional brain networks;
2. The study of these brain features in patients affected by various diseases/disorders using naturalistic paradigm;
3. The study investigating the test-retest reliability of functional connectivity/network derived from naturalistic neuroimaging data.
Cognitive neuroscience has traditionally focused on the design of simple tasks to isolate a particular cognitive domain for investigation. While this form of research has yielded fundamental knowledge on the functional specialization in the brain, its ecological validity remains unclear. Accordingly, a recent development in neuroimaging is to design ecologically valid naturalistic paradigms that approximate real-life scenarios, using naturalistic and dynamic stimuli such as films and music. Comparing to traditional task-based literature, studying neural processing and network interactions during natural, dynamic conditions is imperative to the understanding of human brain function, further offering a powerful tool for clinical research on brain disorders, for their high compliance and low demand.
While naturalistic paradigms are increasingly used to map brain functions in healthy populations or altered brain functions in clinical populations, one challenge is the lack of effective computational models to characterize neural correlates of naturalistic stimuli. To date, various methods have been successfully applied to characterize different aspects of brain activities under naturalistic paradigms, such as General Linear Model (GLM), and Independent Component Analysis (ICA), and Inter-subject Correlation (ISC), etc. However, there are still many challenges in properly modeling the spatial-temporal characteristics and hidden features from these high-dimensional neuroimaging data with multimodal, dynamic naturalistic stimuli.
Recently, data mining and modeling methods have developed in machine learning and artificial intelligence, gaining increasing attention in the computational neuroscience field. These innovative methods can be nicely positioned to exploit big neuroimaging data sets by tackling the complex challenges presented by multimodal contents of the stimuli.
The aim of this Research Topic is to provide an interdisciplinary platform for researchers to investigate the computational methods for modeling neuroimaging data (fMRI, EEG, ECoG, MEG, etc.) that employ the rich, multimodal dynamic stimuli, such as film clips, TV advertisements, news items, and spoken narratives, or that embody relatively unconstrained interactions with other agents, gaming environments, or virtual realities.
We welcome original research articles and review articles that focus on one of the three following aspects, or all:
1. Methodological advances in the analysis of neuroimaging data using naturalistic paradigm (such as fMRI, EEG, MEG, etc) that allow the identification of dynamic functional brain states or functional brain networks;
2. The study of these brain features in patients affected by various diseases/disorders using naturalistic paradigm;
3. The study investigating the test-retest reliability of functional connectivity/network derived from naturalistic neuroimaging data.