| Popis: |
Metabolic rate is not routinely assessed in healthcare for the general population, nor is it a measure commonly recorded for in-patients (incorrect feeding can slow post-operation recovery rate). For the general community, this lack of knowledge prevents the accurate determination of calorific need and is a factor contributing towards the onset of an overweight and an increasingly obese population. In the UK alone, obesity costs the National Health Service a staggering £5 billion annually. In this thesis a novel low-cost hand-held breath analyser is presented in order to measure human energy expenditure (EE). A unique optical CO2 sensor was developed, capable of sampling exhaled breath with a fast response time ~1 s and resilience to a humidity range of ~30 % to near saturated. The device was tested in a laboratory gas testing rig and a detection limit of ~25 ppm CO2 was measured. A low power metal oxide sensor (~100 mW) was developed to detect volatile organic compounds (VOCs) in the breath, for disease detection and investigation of the variation of inter-individual metabolism processes. The device was sensitive to acetone (100 to 300 ppm, which is a biomarker for type-I diabetes). Other VOCs, such as NO2 were tested (10 to 250 ppb). Further work includes investigating the inter-individual variance of metabolism processes, for which the metal oxide sensor would be well-suited. Software was developed to operate the gas testing rig and acquire sensor output data in real-time. An application was written for smartphones to enable EE measurements with the breath analyser, outside of a laboratory environment. Three hand-held analysers were constructed and tested with a trial of 10 subjects. A counterpart (benchmark) unit with medical grade commercial sensors (cost of ~£2500) and hospital respiratory rooms (reference) were included in the trial. The newly developed analysers improved upon the performance of the benchmark system (average EE measurement error +2.4 % compared to +7.9 %). The affordable device offered far greater accuracy than the traditional method often used by practitioners (predictive equations, error +41.4%). It is proposed a set of periodic (hourly) breath measurements could be used to determine daily EE. The EE analyser and associated low-cost sensors developed in this work offer a potential solution to halt the growing cost of an obese population and provide point-of-care health management. |
