Research Areas: Artificial Intelligence, Machine Learning, EEG Signal Processing, Neuroscience, Healthcare AI, Explainable AI

Project Description:
Attention-Deficit/Hyperactivity Disorder (ADHD) is commonly evaluated using behavioral and clinical assessments. This project investigates whether patterns in electroencephalography (EEG) brain signals can provide additional quantitative information useful for studying attention-related neurological characteristics.

Students will process EEG recordings, extract time-domain and frequency-domain characteristics, and analyze brain-wave activity in the delta, theta, alpha, beta, and gamma bands. Particular attention will be given to features associated with attention and cognitive activity, including spectral power, band-power ratios, signal variability, and connectivity-related measures.

Machine-learning models such as Logistic Regression, Support Vector Machines, Random Forest, XGBoost, and neural-network approaches may be compared for classifying EEG patterns. Explainable-AI methods such as SHAP can be used to investigate which EEG features contribute most strongly to model predictions.

Research Questions: Can EEG-derived features distinguish ADHD-related patterns from comparison groups? Which EEG frequency bands and channels provide the most informative features? How does model performance change across subjects? Can explainable AI help identify EEG characteristics associated with model predictions?

Student Activities: EEG preprocessing, artifact removal, filtering, signal segmentation, spectral analysis, feature extraction, machine-learning model development, visualization, statistical analysis, explainable AI, and scientific writing.

Technologies: Python, MNE-Python, NumPy, SciPy, Pandas, Scikit-learn, XGBoost, SHAP, Matplotlib, Jupyter/Google Colab.

Expected Outcomes: An EEG-processing pipeline, comparative ML study, interpretable feature analysis, research poster/presentation, and potential conference or journal paper.