Add to Google Calendar
Date:  Thu, November 09, 2023
Time:  10:00am - 12:00pm
Location:  online only, see below registration info
Speaker:  Mohammad Shadman Ishrak, candidate for PhD, advisor: Dr. Victor Lubecke

Abstract

Continuous-wave (CW) homodyne Doppler radars offer a non-invasive and unobtrusive means to measure vital signs, holding promise for real-time monitoring of heart rate, respiration, and even sleep apnea detection. These radars work by detecting the motion of the chest wall during breathing and heartbeat. The technology, refined through ongoing research, is also being explored for applications in sleep posture recognition and identity authentication, augmenting its potential in healthcare and security sectors. The accuracy of these radars in capturing the subtle movements associated with breathing and heartbeat has even led to their integration into drone-mounted systems for remote monitoring, especially useful in emergency and triage situations.

A critical aspect of utilizing CW Doppler radars lies in the effective demodulation of data to extract meaningful insights. Various methods, including Arctangent Demodulation (AD), Differentiate and Cross Multiply (EDACM), and Single Channel Demodulation (ScD), have been explored to refine the extraction of periodic respiratory motion from raw data. A comparative analysis using a robotic mover simulating fixed frequency sinusoidal motion revealed AD's superior performance, especially for ac-coupled data. As an extension of the research, additional demodulation methods - Modified Divide and Cross Multiply(MDACM), Linear Demodulation(LD), Complex Signal Demodulation(CSD), PB-DCT(Polyphase Basis Discrete Cosine Transform), Quadrature Cosine Transform(QCT)- were tested for computational complexity, waveform distortion and vital signs frequency detection. The demodulation methods were tested on a robotic mover mimicking the chest wall due to cardiorespiratory activity. Further research is planned to produce a multi-method demodulation for more accurate and efficient vital signs monitoring.

The research also extends into the development of algorithms for non-sedentary activity detection, ensuring the precision of data during physical movements. The phase change derived from arctangent demodulation is directly proportional to target displacement and exhibits large displacements during locomotion, which can be used to differentiate between sedentary and non-sedentary motion. The work is extended to differentiating between Extraneous Body Motions(EBM) during sedentary periods. Future work will focus on the recovery ofcardiorespiratory information in the presence of EBM.

Bio

Mohammad Shadman Ishrak is a Ph.D. candidate in Electrical and Computer Engineering at the University of Hawaii at Manoa as of Fall 2021. His selected track is Electrophysics (EP). His current research interests include RF and microwave circuit design, digital signal processing, demodulation methods, and non-contact physiological monitoring.

Online only, register for connection info at https://forms.gle/yeGtuLSFYqgbEJg86