Related Experiment Video
Updated: Jan 18, 2026
09:47
Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
Published on: December 15, 2023
1.7K
Enhancing Alzheimer's Diagnosis with Machine Learning on EEG: A Spectral Feature-Based Comparative Analysis.
Yeliz Senkaya1, Cetin Kurnaz2, Ferdi Ozbilgin3
1Department of Computer Applications, Akkus Vocational School, Ordu University, 52950 Ordu, Türkiye.
Diagnostics (Basel, Switzerland)
|September 13, 2025
Summary
Machine learning enhances Alzheimer's disease diagnosis using electroencephalography (EEG) data. This approach achieved 96.01% accuracy, enabling earlier detection and intervention for improved patient outcomes.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Artificial Intelligence
Background:
- Alzheimer's disease (AD) diagnosis is challenging and often delayed, impacting early intervention effectiveness.
- Electroencephalography (EEG) shows potential for detecting AD-related brain changes, but feature extraction is complex.
- Accelerating AD diagnosis is crucial for managing disease progression due to the lack of a cure.
Purpose of the Study:
- To develop and validate a machine learning (ML) model for enhanced Alzheimer's disease diagnosis using EEG signals.
- To investigate the efficacy of ML algorithms in differentiating AD patients from those with Frontotemporal Dementia (FTD) and healthy controls (HC).
- To identify optimal features from EEG data for accurate AD classification.
Main Methods:
- EEG recordings from 36 AD patients, 23 FTD patients, and 29 HC were analyzed.
- Power Spectral Density (PSD) was computed, and 342 statistical and spectral features were extracted from time-domain and frequency-domain EEG data.
- Support Vector Machines (SVM) and k-Nearest Neighbors (k-NN) classifiers were trained and optimized using Bayesian optimization and feature selection via correlation analyses.
Main Results:
- The Support Vector Machines (SVM) classifier achieved a high diagnostic accuracy of 96.01%.
- The ML approach demonstrated superior performance compared to previously reported methods for AD detection using EEG.
- Feature selection and hyperparameter optimization significantly improved classification accuracy.
Conclusions:
- Machine learning analysis of EEG data presents a promising and effective method for the early diagnosis of Alzheimer's disease.
- This approach facilitates timely clinical interventions, potentially leading to better patient management and outcomes.
- The study highlights the potential of non-invasive EEG combined with advanced ML techniques for neurological disorder diagnostics.
