Epilepsy, affecting 1% of the American population, presents a significant challenge in healthcare and quality of life. This project addresses the complexities of epilepsy diagnosis by introducing a Python-based Graphical User Interface (GUI) for EEG analysis and a novel approach to utilizing semiology for distinguishing between psychogenic non-epileptic seizures (PNES) and epileptic seizures.

The main goal of this work is to streamline the diagnostic process for epilepsy. The GUI serves as a tool that can ease the labor-intensive and error-prone process of manual EEG review, and enables users to seamlessly navigate EEG data, switch between montages, and adjust visualization parameters. By efficiently managing increased amounts of data and user input, the design allows for easy scalability.

The project also investigated the diagnostic potential of semiology. Natural language processing (NLP) techniques were used to differentiate between psychogenic non-epileptic seizures and epileptic seizures. The approach involved analyzing clinician notes, highlighting patterns of similarity by computing cosine similarities. Future work aims to implement machine learning models to predict epilepsy presence to enhance diagnostic accuracy and patient care.