Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

An ADHD Diagnostic Interface Based on EEG Spectrograms and Deep Learning Techniques

Medha Pappula, Syed Muhammad Anwar

TL;DR

This work tackles the challenge of ADHD diagnosis by reducing reliance on behavior-based assessments, which can be biased by gender and misdiagnosis. It combines EEG signals converted into spectrograms via Continuous Wavelet Transform with a ResNet-18 CNN to classify ADHD and to extract informative brain regions. The study identifies frontopolar, parietal, and occipital lobes as key contributors and reports an overall F1 score around 0.90, supporting the development of a three-part cognitive screening interface for school use. The results suggest a feasible, cost-effective pathway for early ADHD screening in educational settings, leveraging inexpensive EEG hardware and DL-based analysis.

Abstract

This paper introduces an innovative approach to Attention-deficit/hyperactivity disorder (ADHD) diagnosis by employing deep learning (DL) techniques on electroencephalography (EEG) signals. This method addresses the limitations of current behavior-based diagnostic methods, which often lead to misdiagnosis and gender bias. By utilizing a publicly available EEG dataset and converting the signals into spectrograms, a Resnet-18 convolutional neural network (CNN) architecture was used to extract features for ADHD classification. The model achieved a high precision, recall, and an overall F1 score of 0.9. Feature extraction highlighted significant brain regions (frontopolar, parietal, and occipital lobes) associated with ADHD. These insights guided the creation of a three-part digital diagnostic system, facilitating cost-effective and accessible ADHD screening, especially in school environments. This system enables earlier and more accurate identification of students at risk for ADHD, providing timely support to enhance their developmental outcomes. This study showcases the potential of integrating EEG analysis with DL to enhance ADHD diagnostics, presenting a viable alternative to traditional methods.

An ADHD Diagnostic Interface Based on EEG Spectrograms and Deep Learning Techniques

TL;DR

This work tackles the challenge of ADHD diagnosis by reducing reliance on behavior-based assessments, which can be biased by gender and misdiagnosis. It combines EEG signals converted into spectrograms via Continuous Wavelet Transform with a ResNet-18 CNN to classify ADHD and to extract informative brain regions. The study identifies frontopolar, parietal, and occipital lobes as key contributors and reports an overall F1 score around 0.90, supporting the development of a three-part cognitive screening interface for school use. The results suggest a feasible, cost-effective pathway for early ADHD screening in educational settings, leveraging inexpensive EEG hardware and DL-based analysis.

Abstract

This paper introduces an innovative approach to Attention-deficit/hyperactivity disorder (ADHD) diagnosis by employing deep learning (DL) techniques on electroencephalography (EEG) signals. This method addresses the limitations of current behavior-based diagnostic methods, which often lead to misdiagnosis and gender bias. By utilizing a publicly available EEG dataset and converting the signals into spectrograms, a Resnet-18 convolutional neural network (CNN) architecture was used to extract features for ADHD classification. The model achieved a high precision, recall, and an overall F1 score of 0.9. Feature extraction highlighted significant brain regions (frontopolar, parietal, and occipital lobes) associated with ADHD. These insights guided the creation of a three-part digital diagnostic system, facilitating cost-effective and accessible ADHD screening, especially in school environments. This system enables earlier and more accurate identification of students at risk for ADHD, providing timely support to enhance their developmental outcomes. This study showcases the potential of integrating EEG analysis with DL to enhance ADHD diagnostics, presenting a viable alternative to traditional methods.

Paper Structure

This paper contains 10 sections, 1 equation, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Proposed Methodology
  3. Electroencephalography (EEG) data preparation
  4. Spectrogram generation
  5. Deep Learning Model
  6. Feature Extraction
  7. Experimental Results
  8. Significant EEG channels
  9. Front-end interface
  10. Conclusion

Figures (4)

  • Figure 1: Our proposed approach for ADHD detection using features extracted from EEG spectrograms and deep learning model.
  • Figure 2: Samples of CWT for children with (left) and without (right) ADHD.
  • Figure 3: Relative importance of each EEG channel in overall ADHD prediction.
  • Figure 4: Visual representation of the ADHD identification tests.