A Novel Deep Convolutional Neural Network Model for Alzheimer's Disease Classification Using Brain MRI.

Autor: Chaimae Ouchicha, Ammor, Ouafae, Meknassi, Mohammed
Zdroj: Automatic Control & Computer Sciences; Jun2022, Vol. 56 Issue 3, p261-271, 11p
Abstrakt: Alzheimer's disease (AD) is a slowly progressive degenerative neurological disease characterized by severe degeneration of cortical neurons. Detecting Alzheimer's is a difficult and time consuming task. Therefore, developing an automated analysis system that is fast, costs less, and is more reliable is essential to help clinicians in the early diagnosis. In this paper, we proposed a deep convolutional neural network (IRDNet) to classify Alzheimer's disease from normal controls (NC) using brain magnetic resonance (MR) images. The proposed IRDNet is based on residual network with dense block to obtain a higher performance over the original residual network and is constructed by using three parallel levels with receptive fields of different sizes to capture global and local features of the inputs. Our model is evaluated and trained on brain MRI images acquired from the Open Access Series of Imaging Studies database (OASIS). The experimental results reveal that our IRDNet achieved an average accuracy of 98.53% after training it for 3.127 seconds for two-class classification (AD vs. normal), which represents a promising classification performance. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index
Externí odkaz: