Measurement and assessment of human energy expenditure and locomotion

Measurements of human energy expenditure (EE) and locomotion have applications in both biomedical research and clinical practice. The present study explored various means of obtaining such measurements.; A room-size indirect human respiratory calorimeter was designed and built for measuring 24-hour energy expenditure (TEE). An ARMA model was found the best description of the system with a stable response and small number of parameters. Several signal processing techniques including frequency analysis, moving average, and system identification were used to significantly improve the system's performance. The system error was controlled within ±2.0%.; The calorimeter is the most accurate method available now to measure TEE. Unfortunately, it is impossible to differentiate the components of TEE, especially the EE caused by a subject's movement. Thus, a large force platform, which was modeled as multiple spring-dashpot systems, was built inside the calorimeter to monitor the subject's motion. This enables us to estimate the energy cost of subject's movement by calculating external mechanical work.; As an application of those hybrid systems, sleeping metabolic rate (SMR) in relation to body mass index (BMI) and body composition was investigated. This study found that SMR decreases as the sleep cycle progresses. This decline was sharper in individuals with a higher BMI (correlation is −0.896).; The calorimetry/force platform systems were shown to measure human energetics efficiently, however, all activities must be confined in a sealed and space-limited room. Information about patterns and energetics of locomotion in free-living subjects still could not be explored by these means. In this work, a novel method, which is able to measure and assess free-living human locomotion, is presented. A portable objective mechanical measurement system was developed to collect the feet-ground contact information step by step. Using temporal measurements of walking and running gaits, artificial neural networks were designed and trained to identify locomotion and predict the speed of walking and running. The average identification rate was more than 98%.; This dissertation, for first time, provides the systematic tools for measuring and assessing human EE and locomotion in both the confined and free-living environments.