•  
  •  
 

Article Type

Article

Abstract

In addition to possible other symptoms, memory loss and impairment are the hallmarks of Demented or Alzheimer’s disease (AD). Despite the fact that dementia is incurable and has a significant negative impact on patients' lives, an early diagnosis can help start the right treatment and prevent additional brain damage. Over the years, machine learning techniques have been used to classify AD; nevertheless, the efficacy of the results depends on the use of multi-step classifiers and manually created features. Thanks to recent advances in deep learning, patterns may now be classified using neural networks' final stage. In order to diagnose dementia early, convolutional neural networks (CNN, VGG16) were utilized in conjunction with magnetic resonance imaging (MRI) to extract features from images of individuals with the condition and classify them. This research focuses on this process (MRI). Gray and white matter MRI image slices were employed as inputs for categorization. The output of deep learning classifiers was combined using group learning techniques to enhance classification after convolutional operations. We assessed the usefulness of our approach in the early detection of this illness with a collection of data from the Demented Disease Neuroimaging Initiative. Our accuracy rate for dementia evaluations was 95.4762%.

Keywords

Demented disease, Deep learning, CNN, AI.

Download

Share

COinS