Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Interpretable feature-transformer framework for cross-subject MCI detection using nonlinear dynamical and graph-theoretic EEG features.

Cognitive neurodynamics·2026

Same author

Detecting Mild Cognitive Impairment to Alzheimer's Disease Progression by fMRI Using Convolutional Neural Network and Long-short Term Memory.

Basic and clinical neuroscience·2026

Same author

Application of artificial intelligence in colonoscopy imaging for polyp analysis-A systematic review.

Computer methods and programs in biomedicine·2026

Same author

An explainable machine learning-based approach to predicting treatment response for neurofeedback in ADHD.

Scientific reports·2025

Same author

Iranian women's perceptions of human papillomavirus and barriers to vaccination: a qualitative study based on the health belief model.

BMC public health·2025

Same author

Classification of Nonalcoholic Fatty Liver Grades using Pre-Trained Convolutional Neural Networks and a Random Forest Classifier on B-Mode Ultrasound Images.

Journal of biomedical physics & engineering·2025

Search research articles

Related Experiment Video

Updated: Feb 24, 2026

Dynamic Inter-subject Functional Connectivity Reveals Moment-to-Moment Brain Network Configurations Driven by Continuous or Communication Paradigms

Dynamic Inter-subject Functional Connectivity Reveals Moment-to-Moment Brain Network Configurations Driven by Continuous or Communication Paradigms

Published on: March 21, 2019

Interpretable Feature-Transformer Framework for Cross-Subject MCI Detection Using Nonlinear Dynamical and

Hadi Azizpour Lindi¹, Reza Shalbaf¹, Ahmad Shalbaf²

¹Cognitive Modeling, Institute for Cognitive Science Studies, Chamran, Pardis, 1658344575, Tehran, Iran.

Research Square

|February 23, 2026

Summary

Early detection of Mild Cognitive Impairment (MCI) is crucial for Alzheimer's disease (AD) prevention. This study shows that combining EEG-derived entropy and graph features with Transformer networks effectively distinguishes MCI from healthy controls.

Keywords:

Alzheimer’s disease Deep Learning EEG EEGNet Entropy Graph Theory MCI Nonlinear Dynamics Resting State SHAP Transformer

More Related Videos

Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms

Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms

Published on: November 1, 2019

Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging

Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging

Published on: June 30, 2018

Related Experiment Videos

Last Updated: Feb 24, 2026

Dynamic Inter-subject Functional Connectivity Reveals Moment-to-Moment Brain Network Configurations Driven by Continuous or Communication Paradigms

Dynamic Inter-subject Functional Connectivity Reveals Moment-to-Moment Brain Network Configurations Driven by Continuous or Communication Paradigms

Published on: March 21, 2019

Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms

Statistical Modelling of Cortical Connectivity Using Non-invasive Electroencephalograms

Published on: November 1, 2019

Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging

Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging

Published on: June 30, 2018

Area of Science:

Neuroscience
Biomedical Engineering
Artificial Intelligence

Background:

Mild Cognitive Impairment (MCI) is a precursor to Alzheimer's disease (AD), necessitating early detection for intervention.
Electroencephalography (EEG) offers a non-invasive method for assessing brain activity relevant to cognitive function.
Distinguishing MCI from healthy controls (HC) is challenging but critical for timely therapeutic strategies.

Purpose of the Study:

To evaluate the efficacy of entropy- and graph-based EEG features in differentiating MCI from HC.
To compare the performance of a Transformer network against an EEGNet model using these engineered features.
To leverage interpretable deep learning for identifying key biomarkers for early MCI detection.

Main Methods:

Utilized resting-state, eyes-closed EEG data from 183 participants (127 HC, 56 MCI).
Extracted nonlinear dynamical measures (entropy, fractal dimension, Lyapunov exponent) and graph-theoretic connectivity features across five frequency bands.
Applied a Transformer network and an EEGNet model to the engineered feature set for classification.

Main Results:

The feature-based Transformer model achieved a high classification accuracy of 97.04% ± 0.72.
The Transformer model outperformed the EEGNet baseline, demonstrating the advantage of attention-based architectures with rich features.
SHAP analysis identified influential nonlinear and connectivity features, along with key EEG channels, enhancing model interpretability.

Conclusions:

Integrating handcrafted EEG features with Transformer networks provides a powerful and interpretable approach for early MCI detection.
This method shows significant potential for developing advanced diagnostic tools to combat Alzheimer's disease progression.
Feature-driven deep learning models offer a promising avenue for understanding and identifying neurodegenerative conditions like MCI.