About this Research Topic
Brain regional activities and inter-regional interactions are changed from one mental state to another mental state or from healthy status to diseased status. This change can be manifested in neuroimaging signals, which are measured from brain. Multiple modal signals have been utilized to probe into brain states or statuses as different signal modalities are of different strengths. For instance, electroencephalogram (EEG) has high temporal resolution, which enables us to explore transient changes in the brain. Whereas, it suffers from the drawback of low spatial resolution. This can be complemented by functional magnetic resonance imaging (fMRI) signal, which is of high spatial resolution. Such complementary signals provide abundant information relevant to human brain. However, it is still challenging to extract intrinsic characteristics from these diverse signals.
Deep learning as one of machine learning methods is usually utilized to learn representations of data by multiple processing layers. The representation is more abstract at a higher layer compared to that at a lower layer. If class labels are used to guide the learning of the deep learning model, the higher layer amplifies differential characteristics between classes and suppresses irrelevant variations. Deep learning has been successfully applied to a number of fields including image retrieval, video processing, speech recognition, audio recognition, artificial intelligence. However, it is relatively limited for the applications of deep learning in neuroimaging signals. This limited developments may be partially due to that the developed deep learning models are not so suitable to neuroimaging signals. Additional effort such as development of proper filtering layer is required to improve the applicability of deep learning model.
This Research Topic aims to collect recent achievements about the developments of deep learning in analyzing, decoding, and understanding neuroimaging signals. The Research Topic will cover any kind of neuroimaging signals, such as EEG, fMRI, and diffusion tensor imaging (DTI) and functional near-infrared spectroscopy (fNIRS), which are measured from human brain. The work focuses on the understanding of neural mechanisms involving in mental states (e.g., mental fatigue and mental workload), brain diseases (e.g., schizophrenia, autism), and aging (e.g., cognitive decline) using deep learning models is particularly welcome. The work focuses on the diagnosis of brain diseases, classification of mental states, and decoding of mental tasks by means of deep learning methods is also welcome. Besides the report of accuracy performance, a further exploration or explanation of feature distribution is preferable, which could deepen our understanding to the brain.
We welcome submissions of both original research and review manuscripts relevant to the topics including, but not limited to:
1. Neuroimaging data analysis using deep learning methods
2. Representation differences between healthy brain and diseased brain using deep learning methods
3. Brain disease diagnosis using deep learning models with the exploration of feature distribution.
4. Mental state/task decoding using deep learning models with the exploration of feature distribution.
5. Improvement of deep learning model for above applications.
Contributors are encouraged to submit an abstract first before the submission of full manuscript.
Deep learning as one of machine learning methods is usually utilized to learn representations of data by multiple processing layers. The representation is more abstract at a higher layer compared to that at a lower layer. If class labels are used to guide the learning of the deep learning model, the higher layer amplifies differential characteristics between classes and suppresses irrelevant variations. Deep learning has been successfully applied to a number of fields including image retrieval, video processing, speech recognition, audio recognition, artificial intelligence. However, it is relatively limited for the applications of deep learning in neuroimaging signals. This limited developments may be partially due to that the developed deep learning models are not so suitable to neuroimaging signals. Additional effort such as development of proper filtering layer is required to improve the applicability of deep learning model.
This Research Topic aims to collect recent achievements about the developments of deep learning in analyzing, decoding, and understanding neuroimaging signals. The Research Topic will cover any kind of neuroimaging signals, such as EEG, fMRI, and diffusion tensor imaging (DTI) and functional near-infrared spectroscopy (fNIRS), which are measured from human brain. The work focuses on the understanding of neural mechanisms involving in mental states (e.g., mental fatigue and mental workload), brain diseases (e.g., schizophrenia, autism), and aging (e.g., cognitive decline) using deep learning models is particularly welcome. The work focuses on the diagnosis of brain diseases, classification of mental states, and decoding of mental tasks by means of deep learning methods is also welcome. Besides the report of accuracy performance, a further exploration or explanation of feature distribution is preferable, which could deepen our understanding to the brain.
We welcome submissions of both original research and review manuscripts relevant to the topics including, but not limited to:
1. Neuroimaging data analysis using deep learning methods
2. Representation differences between healthy brain and diseased brain using deep learning methods
3. Brain disease diagnosis using deep learning models with the exploration of feature distribution.
4. Mental state/task decoding using deep learning models with the exploration of feature distribution.
5. Improvement of deep learning model for above applications.
Contributors are encouraged to submit an abstract first before the submission of full manuscript.
Keywords: Deep Learning, Neuroimaging Data Analysis, Brain Disease Diagnosis, Mental State, Mental Task
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.
