Development of a bed-based nighttime monitoring toolset

Abstract

A movement is occurring within the healthcare field towards evidence-based or preventative care-based medicine, which requires personalized monitoring solutions. For medical technologies to fit within this framework, they need to adapt. Reduced cost of operation, ease-of-use, durability, and acceptance will be critical design considerations that will determine their success. Wearable technologies have shown the capability to monitor physiological signals at a reduced cost, but they require consistent effort from the user. Innovative unobtrusive and autonomous monitoring technologies will be needed to make personalized healthcare a reality. Ballistocardiography, a nearly forgotten field, has reemerged as a promising alternative for unobtrusive physiological monitoring. Heart rate, heart rate variability, respiration rate, movement, and additional hemodynamic features can be estimated from the ballistocardiogram (BCG). This dissertation presents a bed-based nighttime monitoring toolset designed to monitor BCG, respiration, and movement data motivated by the need to quantify the sleep of children with severe disabilities and autism – a capability currently unmet by commercial systems. A review of ballistocardiography instrumentation techniques (Chapter 2) is presented to 1) build an understanding of how the forces generated by the heart are coupled to the measurement apparatus and 2) provide a background of the field. The choice of sensing modalities and acquisition hardware and software for developing the unobtrusive bed-based nighttime monitoring platform is outlined in Chapters 3 and 4. Preliminary results illustrating the system’s ability to track physiological signals are presented in Chapter 5. Analyses were conducted on overnight data acquired from three lower-functioning children with autism (Chapters 6 and 9) who reside at Heartspring, Wichita, KS, where results justified the platform’s multi-sensor architecture and demonstrated the system’s ability to track physiological signals from this sensitive population over many months. Further, this dissertation presents novel BCG signal processing techniques – a signal quality index (Chapter 7) and a preprocessing inverse filter (Chapter 8) that are applicable to any ballistocardiograph. The bed-based nighttime monitoring toolset outlined in this dissertation presents an unobtrusive, autonomous, robust physiological monitoring system that could be used in hospital-based or personalized, home-based medical applications that consist of short or long-term monitoring scenarios.