I designed a flexible patch for continuous, home-based monitoring of both fetal and maternal ECG signals. The device transmits data over Bluetooth to a mobile application, which synchronizes to a cloud server for remote physician review. Signal extraction uses a combination of independent component analysis and template subtraction to isolate the fetal signal from the maternal abdominal recording.

This system received a Proof of Product (POP) grant of $80,000 from UCI Beall Applied Innovation and was validated on 10 pregnant women at UCI Medical Center. Research was featured in IEEE Spectrum.

Fetal ECG Wearable patch Bluetooth Rigid-flex PCB Signal processing iOS / Android

Zebrafish have a unique capacity for cardiac regeneration — they can fully recover from 20% ventricular injury — making them an ideal model organism for studying cardiac disease and drug response. I designed and built a high-throughput ECG recording system capable of simultaneously collecting biosignals from multiple awake zebrafish. The system integrates custom PCB hardware, firmware, a mobile Android application, and perfusion apparatus for sustained recordings.

This platform secured a NIH SBIR Phase 2 grant ($1.6 million) and has been deployed for cardiac regeneration studies and drug screening pipelines.

ECG acquisition Zebrafish model PCB design C/C++ Android MATLAB Altium

I also focused on developing robust and efficient algorithms to extract fetal electrocardiogram (fECG) signals from non-invasive, maternal abdominal recordings. Continuous fetal monitoring is critical for assessing fetal well-being and diagnosing potential congenital heart defects, but extracting clear fetal signals from noisy abdominal data remains a significant technical challenge.

Across my publications, we have explored both advanced filtering techniques and highly optimized deep learning models to make out-of-clinic, continuous fetal monitoring a reality.

Fetal ECG Deep learning Adaptive filtering Wavelet transform MATLAB Signal processing

I investigated multiple electrophysiological and mechanical sensing modalities — including electrocardiography (ECG), photoplethysmography (PPG), speckleplethysmography (SPG), and capacitive pressure sensing — as candidate signal sources for continuous non-invasive blood pressure (CNBP) estimation. This work culminated in a comprehensive methodological review published in IEEE Access (106+ citations) surveying non-invasive blood pressure measurement techniques across sensing paradigms. A subsequent commercialization effort was undertaken through VenaVitals, focusing on pressure sensing technology as the primary acquisition modality.

CNBP ECG PPG SPG Capacitive sensing Signal processing MATLAB

MD-Link is a portable cardiac monitoring system designed for clinical and home use. It features dual mobile applications — one for the patient and one for the physician — synchronized via a cloud server. The system performs automated ECG analysis for early anomaly detection and real-time alerts. Hardware uses a custom multi-layer PCB with ARM Cortex-M microcontroller, and the enclosure is 3D-printed.

MD-Link was selected as 3rd place in the Texas Instruments Innovation Challenge (National Final) and validated in clinical trials with over 200 participants, completing both engineering and design validation phases.

ECG monitoring ARM Cortex-M Android / iOS Cloud sync 3D printing Altium