Based on existing imaging technologies, including functional magnetic resonance imaging (fMRI), structural MRI, and electroencephalogram (EEG), and magnetoencephalogram (MEG), obtaining the images of the brain usually requires a long waiting time and these images often differ from the real brain structures and function. Hence, the accuracy of diagnosing brain diseases may be unsatisfactory. In addition, with the existing deep learning diagnostic algorithms, the effective analysis of acquired images often requires a long preparation time. The conventional brain imaging techniques are usually designed based on the first derivative. In the deep learning diagnostic algorithms, the core is the SGD or Adam algorithm based on the first derivative. Essentially, the algorithms based on the first derivative necessarily converge extremely slowly. In addition, it is exceedingly easy to get stuck in local optima.
As an emerging tool, the fractional calculus is a natural extension of the first derivative, which has achieved great success in the fields of evolutionary computation, system identification, and fluid mechanics. With the help of the non-locality, long-term memory, and weak singularity of the fractional calculus, it not only helps to speed up the convergence speed but also lays a foundation for jumping out of the local optimum. Naturally, applying the framework of fractional calculus to exploring brain imaging and brain disease diagnosis is a novel and meaningful research direction. However, few researchers use the fractional calculus for brain imaging and brain disease diagnosis.
Therefore, the ultimate goal of this Research Topic is to publish high-quality research papers on the intersection of computer science and neuroscience, especially the topics regarding brain imaging and brain disease diagnosis based on the fractional calculus. We aim to explore the practical application value and broad prospects under the system of the fractional calculus, to help promote the rapid and accurate imaging and disease diagnosis of the brain.
Topics of interest include, but are not limited to, the following:
- The fractional calculus algorithms for fast brain imaging.
- Using the fractional calculus algorithms to improve the accuracy of brain imaging.
- Improving the speed of deep learning diagnostic algorithms based on the fractional calculus.
- Designing the fractional calculus algorithms for accurate diagnosis of brain diseases.
- Interpretability of brain imaging and disease diagnosis algorithms based on the fractional calculus.
- Novel brain imaging and disease diagnosis algorithms, including the segmentation of medical images and the diagnosis of neurological diseases through brain imaging, neuroelectrophysiology and other biomarkers based on statistical or machine learning methods.
- Frontier applications and validations in the field of the brain, with a focus on discovering pathologically significant biomarkers associated with neurological diseases, elucidating the physiological mechanisms of neurological diseases, and translating them into clinical diagnosis or treatment.
- Typical application results of neuroscience in other fields, such as law, evolutionary algorithms, and memristors.
Based on existing imaging technologies, including functional magnetic resonance imaging (fMRI), structural MRI, and electroencephalogram (EEG), and magnetoencephalogram (MEG), obtaining the images of the brain usually requires a long waiting time and these images often differ from the real brain structures and function. Hence, the accuracy of diagnosing brain diseases may be unsatisfactory. In addition, with the existing deep learning diagnostic algorithms, the effective analysis of acquired images often requires a long preparation time. The conventional brain imaging techniques are usually designed based on the first derivative. In the deep learning diagnostic algorithms, the core is the SGD or Adam algorithm based on the first derivative. Essentially, the algorithms based on the first derivative necessarily converge extremely slowly. In addition, it is exceedingly easy to get stuck in local optima.
As an emerging tool, the fractional calculus is a natural extension of the first derivative, which has achieved great success in the fields of evolutionary computation, system identification, and fluid mechanics. With the help of the non-locality, long-term memory, and weak singularity of the fractional calculus, it not only helps to speed up the convergence speed but also lays a foundation for jumping out of the local optimum. Naturally, applying the framework of fractional calculus to exploring brain imaging and brain disease diagnosis is a novel and meaningful research direction. However, few researchers use the fractional calculus for brain imaging and brain disease diagnosis.
Therefore, the ultimate goal of this Research Topic is to publish high-quality research papers on the intersection of computer science and neuroscience, especially the topics regarding brain imaging and brain disease diagnosis based on the fractional calculus. We aim to explore the practical application value and broad prospects under the system of the fractional calculus, to help promote the rapid and accurate imaging and disease diagnosis of the brain.
Topics of interest include, but are not limited to, the following:
- The fractional calculus algorithms for fast brain imaging.
- Using the fractional calculus algorithms to improve the accuracy of brain imaging.
- Improving the speed of deep learning diagnostic algorithms based on the fractional calculus.
- Designing the fractional calculus algorithms for accurate diagnosis of brain diseases.
- Interpretability of brain imaging and disease diagnosis algorithms based on the fractional calculus.
- Novel brain imaging and disease diagnosis algorithms, including the segmentation of medical images and the diagnosis of neurological diseases through brain imaging, neuroelectrophysiology and other biomarkers based on statistical or machine learning methods.
- Frontier applications and validations in the field of the brain, with a focus on discovering pathologically significant biomarkers associated with neurological diseases, elucidating the physiological mechanisms of neurological diseases, and translating them into clinical diagnosis or treatment.
- Typical application results of neuroscience in other fields, such as law, evolutionary algorithms, and memristors.