Related Experiment Video
Updated: Nov 5, 2025
05:17
Hybrid PET/MRI Imaging of Alzheimer's Disease Based on 18F-AV-1451
Published on: April 18, 2025
479
Diagnosis of Alzheimer's Disease Severity with fMRI Images Using Robust Multitask Feature Extraction Method and
Morteza Amini1, MirMohsen Pedram2,3, AliReza Moradi4,5
1Department of Cognitive Modeling, Institute for Cognitive Science Studies, Shahid Beheshti University, Tehran, Iran.
Computational and Mathematical Methods in Medicine
|May 19, 2021
Summary
A novel convolutional neural network (CNN) accurately diagnoses Alzheimer's disease severity using fMRI scans. This AI approach surpasses traditional machine learning methods for early and precise detection of Alzheimer's stages.
Area of Science:
- Neuroscience and Artificial Intelligence
- Medical Imaging Analysis
Background:
- Alzheimer's disease (AD) is a progressive neurodegenerative disorder with a long incubation period, necessitating early diagnosis.
- Accurate staging of AD is crucial for timely intervention and patient management.
- Functional magnetic resonance imaging (fMRI) offers insights into brain activity relevant to AD progression.
Purpose of the Study:
- To develop and evaluate a novel convolutional neural network (CNN) architecture for diagnosing Alzheimer's disease severity.
- To investigate the relationship between fMRI data and Mini-Mental State Examination (MMSE) scores for AD classification.
- To compare the performance of the proposed CNN against traditional machine learning algorithms.
Main Methods:
- Classification of Alzheimer's disease severity using machine learning algorithms: K-nearest neighbor (KNN), support vector machine (SVM), decision tree (DT), linear discrimination analysis (LDA), and random forest (RF).
- Development of a novel CNN architecture for AD severity diagnosis.
- Feature extraction utilizing a robust multitask feature learning algorithm.
- Correlation analysis between fMRI images and MMSE scores to categorize severity into low, mild, moderate, and severe.
Main Results:
- The proposed CNN achieved the highest diagnostic accuracy at 96.7%, outperforming all other tested machine learning methods.
- CNN demonstrated high sensitivity across all severity levels: 98.1% (low), 95.2% (mild), 89.0% (moderate), and 87.5% (severe).
- Traditional methods showed varying accuracies, with SVM and RF achieving 85.8% and 85.1%, respectively, while KNN, LDA, and DT performed lower.
Conclusions:
- The novel CNN architecture is a highly effective tool for the automatic diagnosis and staging of Alzheimer's disease.
- This AI-driven approach shows significant potential for improving early detection and management of Alzheimer's disease.
- The findings highlight the capability of CNNs in analyzing complex fMRI data for accurate neurodegenerative disease assessment.
