Alzheimer' disease prediction and classification using CT images through machine learning.

Abstract

Numerous surveys using different techniques have been conducted in recent years to accurately classify Alzheimer's disease (AD). This research emphasized the identification of AD through neuroimaging data. However, it is important to identify symptoms as soon as possible when the disease-modifying medications function best during infection before a permanent cognitive impairment develops. The use of automated algorithms to detect the early symptom of AD to this information was very important. Machine Learning (ML) has been proposed for the evaluation of various image segmentation and database techniques. In addition, Visual Geometry Group (VGG)-16 et Improved Faster Recurrent Convolutional Neural Network (IFRCNN) method developed for the ImageNet database utilizing the mathematical model based on action recognition as a feature extractor for categorization work. Experiments are being conducted on the Alzheimer's Neuroimaging Initiative (ADNI) dataset, and the proposed system achieves the 98.32 % accuracy level (Tab. 6, Fig. 4, Ref. 34). Text in PDF www.elis.sk. Keywords: mild cognitive impairment, deep learning, Alzheimer's disease, expected risk.