Health and activity monitoring by wearables in extreme conditions
Abstract
The project deals with the issue of analysis of signals recorded via wearable devices in extreme conditions which are on the daily basis (e.g. soldiers and firefighters). Wearable devices are perspective because they do not restrict the movement of the person and enable monitoring in extreme conditions. Wearables allow to record biosignals such as ECG, EEG, PPG, and motion data included accelerometric and gyroscopic data. The problem is, that a large amount of artefacts caused by extreme conditions complicate analysis of the signals and decrease potential diagnostic yield and accuracy. Our goal is to develop an algorithm for signal quality estimation in order to further analyse only segments of sufficient quality. According to the quality level, we can provide tailored analysis of the signals to maximize the diagnostic yield and accuracy. Within this project we will also develop new mathematical methods for tailored signal analysis including e.g. full-wave ECG analysis. We will also combine: a) various biosignals to mitigate/compensate for effects of noise and/or missing sensor data and b) biosignals with motion data to improve the accuracy of quality assessment. All the developed algorithms will be able to work real-time (with short-time buffer) in order to enable their incorporation to the compact wearable Mayo Clinic device under development for ONR. All the effort of the project will lead to increase of diagnostic yield of data recorded via wearables during excessive physical load of people such as soldiers and police officers. The results of the project contribute to preventive monitoring of activity in extreme conditions, help to increase the reliability of diagnosing and prevent sudden deaths in duty.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 23, 2019
- Source ID
- N629091912006
Entities
People
- Lukas Smital
Organizations
- Brno University of Technology
- Office of Naval Research
- United States Navy