Developed an automated Sleep Stage Classification Algorithm using a Random Forest model to analyze EEG and EOG channels.
Extracted 9 critical features per 30-second epoch, combining Time-Domain and Frequency-Domain metrics to successfully distinguish REM from N1 sleep stages.
Built a robust data pipeline in Python to extract Machine Learning features like SDNN, RMSSD, and LF/HF Ratio from 120-second dynamic ECG buffers for arrhythmia classification.
Engineered real-time C++ DSP algorithms on bare-metal microcontrollers, utilizing a custom R-peak detection logic with a strict 300ms absolute refractory period to ignore T-waves and EMG artifacts.
ECG Signal Recording
We transform complex physiological data into actionable insights, pushing the boundaries of healthcare technology to enhance patient outcomes worldwide.