Article: Alzheimer's disease classification using hybrid Alex-ResNet-50 model Journal: International Journal of Bioinformatics Research and Applications (IJBRA) 2024 Vol.20 No.6 pp.531 - 555 Abstract: Alzheimer's disease (AD), a leading cause of dementia and mortality, presents a growing concern due to its irreversible progression and the rising costs of care. Early detection is crucial for managing AD, which begins with memory deterioration caused by the damage to neurons involved in cognitive functions. Although incurable, treatments can manage its symptoms. This study introduces a hybrid AlexNet+ResNet-50 model for AD diagnosis, utilising a pre-trained convolutional neural network (CNN) through transfer learning to analyse MRI scans. This method classifies MRI images into Alzheimer's disease (AD), moderate cognitive impairment (MCI), and normal control (NC), enhancing model efficiency without starting from scratch. Incorporating transfer learning allows for refining the CNN to categorise these conditions accurately. Our previous work also explored atlas-based segmentation combined with a U-Net model for segmentation, further supporting our findings. The hybrid model demonstrates superior performance, achieving 94.21% accuracy in identifying AD cases, indicating its potential as a highly effective tool for early AD diagnosis and contributing to efforts in managing the disease's impact. Inderscience Publishers - linking academia, business and industry through research

Title: Alzheimer's disease classification using hybrid Alex-ResNet-50 model

Authors: E. Semmalar; R. Shobarani

Addresses: Department of Computer Applications, Dr. M.G.R. Educational and Research Institute, Chennai, Tamil Nadu, India ' Department of Computer Science and Engineering, Dr. M.G.R. Educational and Research Institute, Chennai, Tamil Nadu, India

Abstract: Alzheimer's disease (AD), a leading cause of dementia and mortality, presents a growing concern due to its irreversible progression and the rising costs of care. Early detection is crucial for managing AD, which begins with memory deterioration caused by the damage to neurons involved in cognitive functions. Although incurable, treatments can manage its symptoms. This study introduces a hybrid AlexNet+ResNet-50 model for AD diagnosis, utilising a pre-trained convolutional neural network (CNN) through transfer learning to analyse MRI scans. This method classifies MRI images into Alzheimer's disease (AD), moderate cognitive impairment (MCI), and normal control (NC), enhancing model efficiency without starting from scratch. Incorporating transfer learning allows for refining the CNN to categorise these conditions accurately. Our previous work also explored atlas-based segmentation combined with a U-Net model for segmentation, further supporting our findings. The hybrid model demonstrates superior performance, achieving 94.21% accuracy in identifying AD cases, indicating its potential as a highly effective tool for early AD diagnosis and contributing to efforts in managing the disease's impact.

Keywords: Alzheimer's disease; AD; segmentation; classification; CNN; AlexNet; U-Net; ResNet-50; moderate cognitive impairment; MCI; normal control; NC; MRI images.

DOI: 10.1504/IJBRA.2024.142548

International Journal of Bioinformatics Research and Applications, 2024 Vol.20 No.6, pp.531 - 555

Received: 12 Jan 2024
Accepted: 22 Mar 2024
Published online: 08 Nov 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Alzheimer's disease classification using hybrid Alex-ResNet-50 model

Keep up-to-date