As the world’s population ages rapidly, dementia is becoming a major global health problem. Although there are drugs, such as cholinesterase inhibitors, used to temporarily improve the symptoms of dementia, no currently available treatments are able to cure it. Therefore, the emphasis is now placed on the prevention of the onset of dementia. To this end, it is important to establish diagnostic technologies for the early detection of cognitive impairment. In order to resolve these issues, this Research Topic welcomes papers that apply machine learning technologies, including ”deep learning”, to the detection of cognitive impairment and dementia.
Deep learning (DL) is a subset of machine learning, that uses deep neural networks and analyzes regularity and relevance from a large amount of complex data, allowing judgments and predictions to be made. During recent years, this branch of data science has progressed in quantum leaps, generating innovation in various fields including medicine. In the medical field, machine learning has been successfully applied in various areas including cancer diagnosis, image-based diagnosis, genetics-based diagnosis and prediction of treatment outcome. Most importantly it has been shown, in many cases, to outperform the diagnostic ability of human experts. However, the application of DL to the diagnosis of dementia and risk assessment of cognitive impairment is relatively limited compared to other medical fields.
This Research Topic aims to encourage clinicians working with dementia patients to deepen their understanding of machine learning, including DL, which is expected to stimulate the discovery of new methods and applications.
We call for Original Research and Reviews on the uses of deep learning in the medical context towards the diagnosis of dementia and risk assessment of cognitive decline, especially for the classification, estimation, analysis and prediction of:
• Imaging data (MRI, PET, CT);
• Blood data;
• Physiological data (EEG, NIRS);
• Data relating to behaviors including facial expressions, voice, etc.
We would like to acknowledge that Dr. Forrest Sheng Bao (Iowa State University, USA) and Dr. Koichi Kurumatani (National Institute of Advanced Industrial Science and Technology, Japan) have acted as coordinators and have contributed to the preparation of the proposal for this Research Topic. 