Brain imaging has been successfully used to generate image-based biomarkers for various neurological and psychiatric disorders, such as Alzheimer’s and related dementias, Parkinson’s disease, stroke, traumatic brain injury, brain tumors, depression, schizophrenia, etc. However, accurate brain image-based diagnosis at the individual level remains elusive, and this applies to the diagnosis of neuropathological diseases as well as clinical syndromes. In recent years, deep learning techniques, due to their ability to learn complex patterns from large amounts of data, have had remarkable success in various fields, such as computer vision and natural language processing. Applying deep learning methods to brain imaging-assisted diagnosis, while promising, is facing challenges such as insufficiently labeled data, difficulty in interpreting diagnosis results, variations in data acquisition in multi-site projects, integration of multimodal data, clinical heterogeneity, etc.
The goal of this research topic is to gather cutting-edge research that showcases the application of deep learning methods in brain imaging for the diagnosis of neurological and psychiatric disorders. We encourage submissions that demonstrate novel approaches to overcome various abovementioned difficulties and achieve more accurate, reliable, generalizable, and interpretable diagnosis of neurological and psychiatric disorders in this field.
We welcome contributions that explore different types of deep learning architectures and their applications to brain imaging data. Some of the topics that we would like to cover in this research topic include but not limited to:
• Deep learning models for detection, classification, and prediction of neurological and psychiatric disorders based on various types of brain imaging data, such as MRI, fMRI, PET, etc
• Integration of multimodal brain imaging data using deep learning techniques
• Transfer learning and domain adaptation in brain imaging analysis
• Unsupervised/Semi-supervised training in brain imaging analysis
• Interpretability of diagnosis results for deep learning models
• Data augmentation, preprocessing, and harmonization techniques for deep learning-based brain imaging analysis
• Model compression techniques for deep models in brain imaging analysis
• Reviews and mini-reviews of recent advances highlighting future directions
Brain imaging has been successfully used to generate image-based biomarkers for various neurological and psychiatric disorders, such as Alzheimer’s and related dementias, Parkinson’s disease, stroke, traumatic brain injury, brain tumors, depression, schizophrenia, etc. However, accurate brain image-based diagnosis at the individual level remains elusive, and this applies to the diagnosis of neuropathological diseases as well as clinical syndromes. In recent years, deep learning techniques, due to their ability to learn complex patterns from large amounts of data, have had remarkable success in various fields, such as computer vision and natural language processing. Applying deep learning methods to brain imaging-assisted diagnosis, while promising, is facing challenges such as insufficiently labeled data, difficulty in interpreting diagnosis results, variations in data acquisition in multi-site projects, integration of multimodal data, clinical heterogeneity, etc.
The goal of this research topic is to gather cutting-edge research that showcases the application of deep learning methods in brain imaging for the diagnosis of neurological and psychiatric disorders. We encourage submissions that demonstrate novel approaches to overcome various abovementioned difficulties and achieve more accurate, reliable, generalizable, and interpretable diagnosis of neurological and psychiatric disorders in this field.
We welcome contributions that explore different types of deep learning architectures and their applications to brain imaging data. Some of the topics that we would like to cover in this research topic include but not limited to:
• Deep learning models for detection, classification, and prediction of neurological and psychiatric disorders based on various types of brain imaging data, such as MRI, fMRI, PET, etc
• Integration of multimodal brain imaging data using deep learning techniques
• Transfer learning and domain adaptation in brain imaging analysis
• Unsupervised/Semi-supervised training in brain imaging analysis
• Interpretability of diagnosis results for deep learning models
• Data augmentation, preprocessing, and harmonization techniques for deep learning-based brain imaging analysis
• Model compression techniques for deep models in brain imaging analysis
• Reviews and mini-reviews of recent advances highlighting future directions
