About
I am a multi-disciplinary researcher focused on advancing digital healthcare leveraging signal processing, deep learning, wearable sensing, and emerging technologies with 8 years of PhD and industry experience.
Academic Focus: My PhD research at Cornell University, under the supervision of Professor Edwin Kan, focused on developing novel radio frequency (RF) sensor systems for assisted living applications, detailed in my thesis on Radio Frequency sensor applications for assisted living. I combined expertise in RF systems, signal processing, machine learning, and biomedical sensors, to create innovative solutions for:
- Continuous cardiopulmonary monitoring (Funding: NIH, DoD CDMRP) Link
- Developed wearable near‑field RF sensor prototype for abnormal respiration monitoring & signal processing algorithms for respiratory biomarkers detection in 2-40 BrPM range.
- Designed and performed an IRB approved human study involving simulation of breathing disorders, apnea, and user attention.
- Collaborated with Cornell Weill Medical Sleep Center to collect patient data for sleep apnea detection.
- Assisted living-based occupant counting & location estimation (Funding: DoE ARPA‑E) Link
- Performed indoor un‑tagged occupant counting using ML for occupants in different postures.
- Developed algorithm for high-resolution 3D occupant image reconstruction for location estimation.
- Developed a 1/6 scale prototype to show real-time demo in ARPA-E summit 2019.
Industry Experience: After my PhD, I developed innovative remote patient monitoring (RPM) solutions at Biofourmis, a Boston-based healthcare startup. Leveraging my expertise in signal processing and deep learning, I:
- Improved the accuracy of respiratory rate estimation using PPG and IMU sensors, resulting in a US Patent submission in October 2023.
- Developed DL-based sleep apnea detection and severity estimation algorithms using SpO2 and PR signals, presented at IEEE EMBC 2022, offering a non-invasive and at-home approach for diagnosis and continuous apnea monitoring.
- Contributed to the core RPM platform by tracking vital alert performance, performing case studies, and responding to escalations, ultimately leading to a reduction in false SpO2 alerts due to noisy signals.
Research Interests: Digital healthcare, Noninvasive physiological monitoring, Wearable sensing, Signal processing, Machine Learning & Artificial Intelligence, Biomedical engineering
I am currently looking for the next exciting research opportunity in the above fields. Please reach out to me at sharmapragya.sp@gmail.com.
Link to my Resume.