An approach to a wrist wearable based Covid-19 prediction system to protect Health Care Professionals.

With healthcare professionals being the frontline warriors in battling the Covid pandemic, their risk of exposure to the virus is extremely high. We present our experience in building a system, aimed at monitoring the physiology of these professionals 24/7, with the hope of providing timely detection of infection and thereby better care. We discuss a machine learning approach and model using signals from a wrist wearable device to predict infection at a very early stage. In a double-blind test on a small group of patients, our model could successfully identify the infected versus non-infected cases with near 100% accuracy. We also discuss some of the challenges we faced, both technical and non-technical, to get this system off the ground as well as offer some suggestions that could help tackle these hurdles.

Person and Stressor Independent Generic Model for Stress Detection Using GSR.

Stress detection is a widely researched topic and is important for overall well-being of an individual. Several approaches are used for prediction/classification of stress. Most of these approaches perform well for subject and activity specific scenarios as stress is highly subjective. So, it is difficult to create a generic model for stress prediction. Here, we have proposed an approach for creating a generic stress prediction model by utilizing knowledge from three different datasets. Proposed model has been validated using two open datasets as well as on a set of data collected in our lab. Results show that the proposed generic model performs well across studies conducted independently and hence can be used for monitoring stress in real life scenarios and to create mass-market stress prediction products.

Effect of Mental Workload on Breathing Pattern and Heart Rate for a Working Memory Task: A Pilot Study.

Mental workload or cognitive load is the total amount of mental resources required while doing a task. Apart from qualitative measures, various physiological signals are being used for assessment of mental workload. However, very limited research has been done on assessment of cognitive load from respiratory signals. In the present study, we have tried to analyze the cognitive load mainly based on respiratory features. n-back memory test has been modified to impart low and high cognitive load. The peripheral blood volume signal (PPG) collected while executing the task is used to reconstruct the breathing pattern signal. A number of morphological as well as statistical features are calculated from this reconstructed signal. Finally a classifier is used for classifying the low and high cognitive load. Results show that a classification accuracy of 76.8% is obtained while using respiratory features only. A maximum accuracy of 81.80% is obtained if we combine time domain PPG features with respiratory features. The features finally selected can also be used to study the habituation effect.