Internet of Medical Things–based decision system for automated classification of Alzheimer’s using three-dimensional views of magnetic resonance imaging scans

Abstract

Internet of Medical Things is a smart provision of medical services to patients interacting with the doctors in harmony to uplift healthcare facilities. It enables the automated diagnosis of diseases for patients in remote areas. Alzheimer’s disease is one of the most chronic diseases and the main cause of dementia in human beings. Dementia affects the patient by a process of gradual degeneration of the human brain and results in an inability to perform daily routine tasks and actions. An automated system needs to be developed, to classify the subject with dementia and to determine the prodromal stage of dementia. Considering such requirement, a fully automated classification system is proposed. The proposed algorithm works on the hybrid feature vector combining the textural, statistical, and shape features extracted from three-dimensional views. The feature length is reduced using principal component analysis and relevant features are extracted for classification. The proposed algorithm is tested for both binary and multi-class problems. The method achieves the average precision of 99.2% and 99.02% for binary and multi-class classifications, respectively. The results outperform the existing methods. The algorithm showed accurate results with the average computational time of 0.05 s per magnetic resonance imaging scan.