Neuroimaging is an interdisciplinary field of physics, engineering, computer vision, mathematics, data science, and medicine, which focuses on the computational analysis of acquiring and interpreting images from the brain. The computational technology in neuroimaging can be categorized into three parts. The first part is imaging, which utilizes various brain imaging techniques, such as MRI, CT, EEG, MEG, PET, to image the structure, function, or pharmacology of the nervous system. Mathematical and computational models play essential roles for image acquisition, including reconstruction, medical physics, hardware and software optimization, spatial-temporal alignment etc. The second part is image processing, to extract quantitative information or metrics from the acquired brain images. Representative image processing approaches are preprocessing, registration, segmentation, surface reconstruction, etc. The third part is data analysis, which investigates and understands the hidden regularities behind the metrics extracted by image processing. The common data analysis approaches include statistical analysis, visualization, modality specific computing, etc.
In the past decade, advancements in data sharing and robust processing have made available considerable quantities of brain images all over the world, which are reshaping the neuroimaging research field from small-scale to large-scale. However, traditional neuroimaging analysis techniques can be inadequate to overcome the new challenges in big data, including the robustness of algorithms, inter-subject variabilities, computational resources, etc. Herein, there is an increasing demand to develop and deploy advanced machine learning algorithms in neuroimaging as well as its clinical applications, including, but not limited to, deep learning in neuroimaging computing, model-based neuroimaging analysis, big data in neuroimaging, and imaging processing platform and informatics.
This Research Topic seeks original contributions to advance neuroimaging using machine learning. More specifically, this collection of articles is intended to seek original contributions for neuroimaging acquisition, quantitative biomarkers computation, multi-modality data analysis, imaging informatics, and prognosis and treatment response analysis, with advanced machine learning algorithms.
Topics of this collection include, but are not limited to, machine learning methods with their applications in:
• Computer-assisted neuroimaging techniques, such as MRI, CT, EEG, MEG, PET etc.
• Machine learning algorithms for structural and functional image analysis, image reconstruction, computer-aided intervention, multi-center data analysis. The topics cover a broad spectrum of machine learning algorithms, such as statistical learning, model-based machine learning, expert systems, deep neural networks, reinforcement learning, and generative adversarial learning.
• Machine learning algorithms for multi-modal data analysis. The examples include multi-modal machine learning algorithms that integrate neuroimaging data with the genetic, clinical report, electrical medical record etc.
• Application studies of using quantitative neuroimaging analysis with machine learning, for, but not limited to, neuro-degeneration, neurological disorders, physiatric disorders, neuro-development, brain lesions, treatment evaluation, prognosis, clinical and cognitive neuroscience.
Dr. D. Jin is working as Senior Scientist for the company PAII Inc.
Dr. D. Guo is working as Senior Research Scientist for the company PAII Inc.
The other Topic Editors declare no competing interests with regards to the Research Topic.
Neuroimaging is an interdisciplinary field of physics, engineering, computer vision, mathematics, data science, and medicine, which focuses on the computational analysis of acquiring and interpreting images from the brain. The computational technology in neuroimaging can be categorized into three parts. The first part is imaging, which utilizes various brain imaging techniques, such as MRI, CT, EEG, MEG, PET, to image the structure, function, or pharmacology of the nervous system. Mathematical and computational models play essential roles for image acquisition, including reconstruction, medical physics, hardware and software optimization, spatial-temporal alignment etc. The second part is image processing, to extract quantitative information or metrics from the acquired brain images. Representative image processing approaches are preprocessing, registration, segmentation, surface reconstruction, etc. The third part is data analysis, which investigates and understands the hidden regularities behind the metrics extracted by image processing. The common data analysis approaches include statistical analysis, visualization, modality specific computing, etc.
In the past decade, advancements in data sharing and robust processing have made available considerable quantities of brain images all over the world, which are reshaping the neuroimaging research field from small-scale to large-scale. However, traditional neuroimaging analysis techniques can be inadequate to overcome the new challenges in big data, including the robustness of algorithms, inter-subject variabilities, computational resources, etc. Herein, there is an increasing demand to develop and deploy advanced machine learning algorithms in neuroimaging as well as its clinical applications, including, but not limited to, deep learning in neuroimaging computing, model-based neuroimaging analysis, big data in neuroimaging, and imaging processing platform and informatics.
This Research Topic seeks original contributions to advance neuroimaging using machine learning. More specifically, this collection of articles is intended to seek original contributions for neuroimaging acquisition, quantitative biomarkers computation, multi-modality data analysis, imaging informatics, and prognosis and treatment response analysis, with advanced machine learning algorithms.
Topics of this collection include, but are not limited to, machine learning methods with their applications in:
• Computer-assisted neuroimaging techniques, such as MRI, CT, EEG, MEG, PET etc.
• Machine learning algorithms for structural and functional image analysis, image reconstruction, computer-aided intervention, multi-center data analysis. The topics cover a broad spectrum of machine learning algorithms, such as statistical learning, model-based machine learning, expert systems, deep neural networks, reinforcement learning, and generative adversarial learning.
• Machine learning algorithms for multi-modal data analysis. The examples include multi-modal machine learning algorithms that integrate neuroimaging data with the genetic, clinical report, electrical medical record etc.
• Application studies of using quantitative neuroimaging analysis with machine learning, for, but not limited to, neuro-degeneration, neurological disorders, physiatric disorders, neuro-development, brain lesions, treatment evaluation, prognosis, clinical and cognitive neuroscience.
Dr. D. Jin is working as Senior Scientist for the company PAII Inc.
Dr. D. Guo is working as Senior Research Scientist for the company PAII Inc.
The other Topic Editors declare no competing interests with regards to the Research Topic.
