About this Research Topic
Brain imaging is a non-invasive approach to map the structure and function of the brain. Commonly used brain imaging methods include magnetic resonance imaging, electrical and magnetic recordings, positron emission tomography, and optical imaging techniques. In the last decade, brain imaging methods combined with artificial neural networks, fuzzy systems, and evolutionary algorithms, amongst others, have been used in computer-aided diagnosis for brain diseases, such as EEG-based seizure classification, and MRI-based brain image segmentation. However, challenges still exist in various brain imaging data processing practices due to missing data, un-labeling data, and data sparsity.
Emerging technologies such as transfer learning (TL), deep learning (DL), multi-view learning (MVL), multi-task learning (MTL), sparse learning (SL), and active learning (AL) have provided new opportunities for developing advanced computational intelligence models which can be used in segmentation, detection, classification, recognition, and prediction of brain imaging data processing. A growing body of research has demonstrated that the new computational intelligence methods, combined with transfer learning, deep learning and multi-task learning can be used to address missing data, un-labeling data, and data sparsity issues in imaging data processing. For example, the newly developed data-driven fuzzy system modeling methods, such as multi-view fuzzy system, multi-task fuzzy system and transfer fuzzy system utilizing MVL, MTL and TL have been applied in analyzing imaging data under various data-deficiency scenes, such as seizure classifications under data missing scene, brain image segmentation under un-labeling data scene, and synthetic CT generation under data sparsity scene, which provides strong support for more accurate diagnoses.
In this research topic, we aim to bring together researchers from computational intelligence and neuroscience to stimulate collaboration. We are particularly interested in the recent paradigm of deep learning and transfer learning in brain imaging data processing dedicated to analysis, diagnosis, and modeling of the neural mechanisms of brain functions.
The sub-topics for this Research Topic include, but are not limited to:
• Supervised methods, such as support vector machine, random forest and K-nearest neighbor
• Unsupervised methods, such as clustering, self-training algorithm and graph-based method
• Semi-supervised method, such as semi-supervised support vector machine and label propagation algorithm
• Probability methods, such as the Monte Carlo method, Bayesian method and conditional random field
• Fuzzy logics, such as fuzzy rough set, fuzzy control and fuzzy system
• Swarm intelligence, such as ant colony optimization, particle swarm optimization and artificial bee colony
• Other formalisms, such as expert system, mathematical analysis and intelligence agent
Applied to the following biomedical imaging technologies:
• X-ray radiography
• Endoscopy
• Ultrasound, Ultrasonography
• CT
• PET, SPECT
• Magnetic Resonance Imaging
• EEG, MEG and EKG as combined with imaging tools
Emerging technologies such as transfer learning (TL), deep learning (DL), multi-view learning (MVL), multi-task learning (MTL), sparse learning (SL), and active learning (AL) have provided new opportunities for developing advanced computational intelligence models which can be used in segmentation, detection, classification, recognition, and prediction of brain imaging data processing. A growing body of research has demonstrated that the new computational intelligence methods, combined with transfer learning, deep learning and multi-task learning can be used to address missing data, un-labeling data, and data sparsity issues in imaging data processing. For example, the newly developed data-driven fuzzy system modeling methods, such as multi-view fuzzy system, multi-task fuzzy system and transfer fuzzy system utilizing MVL, MTL and TL have been applied in analyzing imaging data under various data-deficiency scenes, such as seizure classifications under data missing scene, brain image segmentation under un-labeling data scene, and synthetic CT generation under data sparsity scene, which provides strong support for more accurate diagnoses.
In this research topic, we aim to bring together researchers from computational intelligence and neuroscience to stimulate collaboration. We are particularly interested in the recent paradigm of deep learning and transfer learning in brain imaging data processing dedicated to analysis, diagnosis, and modeling of the neural mechanisms of brain functions.
The sub-topics for this Research Topic include, but are not limited to:
• Supervised methods, such as support vector machine, random forest and K-nearest neighbor
• Unsupervised methods, such as clustering, self-training algorithm and graph-based method
• Semi-supervised method, such as semi-supervised support vector machine and label propagation algorithm
• Probability methods, such as the Monte Carlo method, Bayesian method and conditional random field
• Fuzzy logics, such as fuzzy rough set, fuzzy control and fuzzy system
• Swarm intelligence, such as ant colony optimization, particle swarm optimization and artificial bee colony
• Other formalisms, such as expert system, mathematical analysis and intelligence agent
Applied to the following biomedical imaging technologies:
• X-ray radiography
• Endoscopy
• Ultrasound, Ultrasonography
• CT
• PET, SPECT
• Magnetic Resonance Imaging
• EEG, MEG and EKG as combined with imaging tools
Keywords: Computational Intelligence, Brain imaging processing, Deep learning, Transfer learning
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.
