Coherent ultra-wideband radar for biosensing in dementia

Recent years have seen a wide range of digital and wearable sensing technologies for medical applications. A key challenge remains the compliance of wearable devices that require effort or input from the user. This has been the main reason the impact of wearable device has been limited. A technology with no burden to the user would be thus highly desirable. Recent advances in low power radar-on-chip have enabled new opportunities for miniature, low cost systems deployed in indoor applications. This research will explore the use of coherent ultra-wideband radar for sensing human behaviour and physiology. Sensing targets that will be considered include: respiration, heart rate, heart rate variability, position in two-dimensional space, gait, and movement speed. These parameters will be detected as part of a smart-home platform for the care of people living with dementia, away from hospital premises. This will provide a powerful, unobtrusive tool for sensing gait in rehabilitation, detecting falls, monitoring vital sign variability and sleep activity, observing movement patterns, etc. Such a technology would additionally avoid privacy concerns that exist with camera or imaging technologies. Specific research objectives include: (1) to develop real-time methods for extracting high level information from radar data based on advanced digital signal processing and machine learning techniques; (2) to assess performance against gold standards through validation trials; (3) to investigate real-world scenarios for deployment considering different space configurations, artefact and noise sources; (4) to further advance methods for improving robustness in real-world scenarios; (5) to demonstrate utility of methods in the intended end use. This research aligns with the following UKRI/EPSRC research themes: "Assistive technology, rehabilitation and musculoskeletal biomechanics", "Clinical technologies", "RF and microwave devices", "Sensors and instrumentation".