Research / Final-year project
Abnormal EEG Classification
Two-year research collaboration with Command Hospital, Air Force Bangalore, focused on automated abnormal EEG classification using deep learning and signal-processing techniques.
- Built EEGNet-based pipelines for normal vs abnormal EEG detection using the TUH EEG Corpus.
- Developed a multi-input framework combining raw EEG with handcrafted spectral and non-linear features.
- Implemented MNE-Python preprocessing, memory-efficient incremental training, ensemble prediction, and a Streamlit visualization interface.
PythonEEGNetMNE-PythonDeep LearningSignal ProcessingStreamlit