About me

I am a R&D Signal Processing and Data Science Engineer at Myant Inc. (Toronto) working on several projects including sleep staging, modelling long-term health using overnight data, signal quality assessment of wearables and blood pressure monitoring. Before starting my full-time position, I worked at Myant as a MITACS postdoctoral researcher working on psychophysiological modelling on smart textiles studying ways of improve sleep staging and in-the-wild fatigue detection. I received my B. Tech. and M. Tech. from Indian Institute of Technology, Kharagpur in Engineering Product Design and Manufacturing with specialization in Industrial Electronics in 2016.. Following this, I completed my PhD in Biomedical Signal Processing at Institut National de la Recherche Scientifique (INRS), University of Quebec, Montreal under Prof. Tiago H. Falk (MuSAE Lab) in July 2021. (CV)

For my Master's thesis, I studied the effect of mental fatigue on EEG signals. I used functional connectivity analysis as well as other EEG spectral features for mental fatigue detection. We found significant trends in the different EEG bands with increasing mental fatigue. Further, we made use of noise-robust motif complexity analysis and showed it to be a reliable metric for fatigue detection.

For my PhD research, I shifted my focus on signal processing for wearable devices in order to develop noise-robust physiologically-motivated features for prediction of mental workload, stress and anxiety. During this time, I explored various non-linear and motif based features for EEG based affect recognition as well as Heart Rate Variability (HRV) features for in-the-wild stress assessment recorded using chest bands and smart shirts.

Currently at Myant, I am exploring the use of physiological signals recorded from smart textile for assessment of various physiological constructs (stress, anxiety and mental fatigue) in real-world conditions. Additionally, I am working on improving sleep staging algorithms by making use of improved HRV features along with innovations in machine learning. A part of my work is also dedicated to development of quality assessment metrics of various biosignals for making important design decisions for smart textile development. More recently, I have been involved in developing continuous cuff-less blood pressure detection techniques using physiological signals.

During my post-doctoral position at Myant, I was also involved in a part-time collaboration with Jewish General Hospital, McGill University as a machine learning and AI researcher under Dr. Jonathan Afilalo for development of interpretable deep learning models for frailty detection. We tested out different input representations of the 12-lead ECG alongside the use of quality metrics for helping detect level of frailty in subjects.

During my free time, I like to stay active with running, hiking and bouldering. I also enjoy attending stand-up comedy shows and playing board games.