Abstract:Attention deficit hyperactivity disorder (ADHD) is one of the common neurodevelopmental disorders in children. This paper presents an automated approach for ADHD detection using the proposed entropy difference (EnD)- based encephalogram (EEG) channel selection approach. In the proposed approach, we selected the most significant EEG channels for the accurate identification of ADHD using an EnD-based channel selection approach. Secondly, a set of features is extracted from the selected channels and fed to a classifier. To verify the effectiveness of the channels selected, we explored three sets of features and classifiers. More specifically, we explored discrete wavelet transform (DWT), empirical mode decomposition (EMD) and symmetrically-weighted local binary pattern (SLBP)-based features. To perform automated classification, we have used k-nearest neighbor (k-NN), Ensemble classifier, and support vectors machine (SVM) classifiers. Our proposed approach yielded the highest accuracy of 99.29% using the public database. In addition, the proposed EnD-based channel selection has consistently provided better classification accuracies than the entropy-based channel selection approach. Also, the developed method

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve This paper aims to address the issue of automatic detection of Attention Deficit Hyperactivity Disorder (ADHD) in children. ADHD is a common neurodevelopmental disorder characterized by inattention, impulsive behavior, and hyperactivity. Currently, the diagnosis of ADHD mainly relies on symptom analysis conducted by clinicians based on the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV and DSM-V), discussions with parents, and questionnaires. However, this subjective assessment method has certain inconsistencies. Therefore, this paper proposes an Electroencephalogram (EEG) channel selection method based on Entropy Difference (EnD) for the automatic detection of ADHD. ### Main Contributions 1. **Introduction of a Channel Selection Method Based on Entropy Difference (EnD) Values**: - By calculating the entropy difference value of each EEG channel between ADHD patients and healthy controls, the most meaningful channels for ADHD detection are selected. 2. **Development of an Automatic ADHD Detection Method**: - Using the important channels selected by the EnD-based channel selection method, features are extracted and input into a classifier to achieve automatic ADHD detection. 3. **Validation of the Proposed Method's Effectiveness**: - Experimental validation was conducted using three different feature sets and three machine learning algorithms (Support Vector Machine SVM, k-Nearest Neighbors k-NN, and Ensemble Learning ENS). The results show that the proposed method achieved a maximum accuracy of 99.29% on a public database, outperforming existing automatic ADHD detection methods. ### Method Overview 1. **Dataset**: - A dataset from IEEE Dataport was used, containing 121 children (ages 7-12), of which 61 were diagnosed with ADHD and 60 were healthy controls. 2. **Channel Selection**: - **Entropy-Based Channel Selection**: Selects the most relevant channels based on the entropy value of the channels. - **Entropy Difference (EnD)-Based Channel Selection**: Calculates the entropy difference value of each channel between the ADHD group and the healthy control group, selecting the channels with the largest entropy difference. 3. **Feature Extraction**: - **Discrete Wavelet Transform (DWT)**: Extracts wavelet coefficients from the EEG signals of each channel and calculates statistical features such as Shannon entropy, mean, variance, and energy. - **Empirical Mode Decomposition (EMD)**: Decomposes the EEG signals into multiple Intrinsic Mode Functions (IMFs) and extracts corresponding statistical features. - **Symmetric Weighted Local Binary Pattern (SLBP)**: Extracts time-domain features to capture local variations in the signals. 4. **Feature Selection**: - Uses the Chi-square test to select the most discriminative features. 5. **Classifier**: - Uses three classifiers for classification: Support Vector Machine (SVM), k-Nearest Neighbors (k-NN), and Ensemble Learning (ENS). ### Experimental Results 1. **70:30 Split Ratio**: - The dataset was split into 70% training set and 30% test set. The results show that the EnD-based channel selection method performed best across all feature extraction and classification methods. 2. **70:30 Random Split Evaluation**: - Repeated 10 times with different random splits, averaging the results to obtain more robust performance metrics. 3. **Ten-Fold Cross-Validation**: - The dataset was divided into 10 folds, with each fold serving as a test set once, and the remaining nine folds as the training set, repeated 10 times. The results show that the EnD-based channel selection method performed excellently across all validation strategies. ### Conclusion The proposed EEG channel selection method based on Entropy Difference (EnD) performs excellently in the automatic detection of ADHD, not only improving classification accuracy but also reducing the complexity of the feature vector, achieving more efficient and effective ADHD detection.

Novel entropy difference-based EEG channel selection technique for automated detection of ADHD

Detection of ADHD From EEG Signals Using Different Entropy Measures and ANN

Automated classification of attention deficit hyperactivity disorder and conduct disorder using entropy features with ECG signals

Topological Feature Search Method for Multichannel EEG: Application in ADHD classification

VHERS: A Novel Variational Mode Decomposition and Hilbert Transform-Based EEG Rhythm Separation for Automatic ADHD Detection

A novel approach to identify the brain regions that best classify ADHD by means of EEG and deep learning

Classification of attention deficit hyperactivity disorder using machine learning on an EEG dataset

Convolutional neural network framework for EEG-based ADHD diagnosis in children

Electroencephalogram (EEG) based prediction of attention deficit hyperactivity disorder (ADHD) using machine learning.

Comparative analysis of electroencephalogram (EEG) data gathered from the frontal region with other brain regions affected by attention deficit hyperactivity disorder (ADHD) through multiresolution analysis and machine learning techniques

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

RESNET34 with Synchrosqueezing Transform for ADHD Disorder Detection Using EEG Signals

ADHD-AID: Aiding Tool for Detecting Children's Attention Deficit Hyperactivity Disorder via EEG-Based Multi-Resolution Analysis and Feature Selection

Detection of ADHD from EEG signals using new hybrid decomposition and deep learning techniques

Detection of ADHD Disorder in Children Using Layer-Wise Relevance Propagation and Convolutional Neural Network: An EEG Analysis

The Hybrid Deep Learning Model for Identification of Attention-Deficit/Hyperactivity Disorder Using EEG

Gabor filter-based statistical features for ADHD detection

Classification of the Children with ADHD and Healthy Children Based on the Directed Phase Transfer Entropy of EEG Signals

Applicable features of electroencephalogram for ADHD diagnosis

A Novel Brain Network Analysis Method for Pediatric ADHD Using RFE-GA Feature Selection Strategy

ADHD/CD-NET: automated EEG-based characterization of ADHD and CD using explainable deep neural network technique