Abstract:Background: EEG signals are commonly used in ADHD diagnosis, but they are often affected by noise and artifacts. Effective preprocessing and segmentation methods can significantly enhance the accuracy and reliability of ADHD classification. Methods: We applied filtering, ASR, and ICA preprocessing techniques to EEG data from children with ADHD and neurotypical controls. The EEG recordings were segmented, and features were extracted and selected based on statistical significance. Classification was performed using various EEG segments and channels with Machine Learning models (SVM, KNN, and XGBoost) to identify the most effective combinations for accurate ADHD diagnosis. Results: Our findings show that models trained on later EEG segments achieved significantly higher accuracy, indicating the potential role of cognitive fatigue in distinguishing ADHD. The highest classification accuracy (86.1%) was achieved using data from the P3, P4, and C3 channels, with key features such as Kurtosis, Katz fractal dimension, and power spectrums in the Delta, Theta, and Alpha bands contributing to the results. Conclusion: This study highlights the importance of preprocessing and segmentation in improving the reliability of ADHD diagnosis through EEG. The results suggest that further research on cognitive fatigue and segmentation could enhance diagnostic accuracy in ADHD patients.

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve This paper aims to improve the accuracy of Attention Deficit Hyperactivity Disorder (ADHD) diagnosis by enhancing the preprocessing and time segmentation methods of EEG (Electroencephalogram) signals. Specifically, the researchers focus on the following issues: 1. **Impact of Preprocessing on Classification**: - How do different EEG signal preprocessing methods affect the classification of individuals with ADHD? - Does the choice of preprocessing method have a decisive impact on the results? - Is there a significant difference in this choice? 2. **Importance of Time Segmentation**: - Does segmenting EEG recordings into different time periods help identify important moments or patterns, thereby improving classification performance? - Particularly in later time periods, are there changes in ADHD characteristics due to cognitive fatigue? 3. **Identification of Key Features and Channels**: - Which EEG features and channels are most helpful for ADHD classification? - By using transparent and interpretable models, ensure that clinicians can trust and understand the diagnostic results of these models. ### Main Contributions - **Importance of Preprocessing**: The study shows that appropriate preprocessing (such as filtering, ASR, and ICA) is crucial for ensuring reliable and unbiased results, avoiding artificially high accuracy due to unclean data. - **New Time Segmentation Method**: A new time segmentation method is introduced, finding that later time periods of EEG recordings provide higher classification accuracy, offering new insights into attention fatigue in ADHD patients. - **Key Features and Channels**: Identified key EEG channels (P3 and P4) and important features (such as kurtosis, Katz fractal dimension, and power spectra of Delta, Theta, and Alpha bands), which are most relevant for ADHD classification. - **High Classification Accuracy**: Achieved up to 86% classification accuracy using data from only three channels, simplifying the data collection process and reducing computational complexity. - **Openness and Reproducibility**: Shared the complete source code, emphasizing openness and reproducibility, promoting further research in this field. ### Conclusion This study systematically evaluates different preprocessing and time segmentation methods, improving the reliability of ADHD diagnosis. The results not only provide new technical means for ADHD diagnosis but also lay the foundation for future related research.

Refining ADHD diagnosis with EEG: The impact of preprocessing and temporal segmentation on classification accuracy

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

Attention deficit and hyperactivity disorder classification with EEG and 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

RESNET34 with Synchrosqueezing Transform for ADHD Disorder Detection Using EEG Signals

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

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

Diagnosis of attention deficit hyperactivity disorder: A deep learning approach

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

Exploring role of prefrontal cortex region of brain in children having ADHD with machine learning: Implications and insights

Assisting Schizophrenia Diagnosis Using Clinical Electroencephalography and Interpretable Graph Neural Networks: a Real-World and Cross-Site Study

EEG data processing in ADHD diagnosis and neurofeedback

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

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

Applicable features of electroencephalogram for ADHD diagnosis

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

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

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

Efficient Deep Learning Approach for Diagnosis of Attention-Deficit/Hyperactivity Disorder in Children Based on EEG Signals

Improved ADHD Diagnosis Using EEG Connectivity and Deep Learning through Combining Pearson Correlation Coefficient and Phase-Locking Value

A deep learning approach with event-related spectral EEG data in attentional deficit hyperactivity disorder