| Millions of people worldwide are afflicted by movement disorder (tremor, gait impairment, etc.). Symptoms of these mobility affecting conditions are often debilitating and demoralizing, sometimes leading to the loss of independence or even death. Efforts to study, diagnose, and treat movement disorders are complicated by a dearth of high quality, motion-based health information. Currently, in-clinic observation and patient self-reports are standards of practice, but both suffer from numerous limitations (e.g. inaccuracy, imprecision, and inconvenience).;A confluence of advancements in diverse areas of research, including device integration, energy storage, sensor technology, and wireless communications, have facilitated the creation of body area sensor networks (BASNs), which are remediating deficiencies in movement disorder care. BASNs consist of networked body area sensor nodes that capture and process continuous, quantitative, accurate, and precise data. The design of such technology is not without its challenges, however, as application requirements increasingly demand smaller (including battery size), more wearable (wirelessly enabled) form factors and longer battery life, which are fundamentally competing energy-centric objectives.;In this dissertation, a holistic and cyclic research approach is applied to design, develop, deploy, and optimize wearable, wireless BASN technology (named TEMPO) for movement disorder applications. The work features three clinical case studies of motion-based health assessment using TEMPO, numerous application-focused signal processing methods that extract clinically-significant information from recorded data, and system optimization techniques formulated to improve BASN performance in healthcare applications. A novel energy-fidelity evaluation framework is presented, which facilitates an outcome-oriented analysis of tradeoffs (e.g. battery life, information quality) resulting from the reduction of wirelessly transmitted data via on-node compression and signal processing. Finally, the benefits of dynamically managed data rate reduction are demonstrated.;The research approach is differentiated by its strict adherence to an authentic medical context, and it is now recognized as a best practice and model of excellence by peers. Consequently, this work fosters significant practical impact (i.e. TEMPO, deployed in numerous healthcare institutions, enables new avenues of motion-based health assessment previously not possible) and offers fundamental research contributions to the body area sensor network community (i.e. in the areas of signal and information processing, and energy-fidelity optimization). |
