| With the development of the social economy and the improvement of thepeople’s living standard, the proportion of overweight or obese people continues torise. Obesity is often accompanied by a variety of chronic diseases, which seriouslyaffects the health and quality of life. Developing a portable device to measure thesubcutaneous adipose tissue thickness (SATT) will meet people’s self-care andphysical health needs.The software and hardware of the portable SATT measuring instrumentpreviously developed were improved based on a series of basic experiments. Humanexperiments were conducted to build fat thickness predictive models. The paperincludes the follows:(1) The basic research on measuring SATT with near infrared of. Accordingto the biological characteristics of human subcutaneous tissue and the principle ofnear infrared measuring subcutaneous fat thickness, we completed a series of basicexperiment of SATT measurement. The light-emitting diode (LED) which wassensitive to SATT was determined.(2) Portable SATT measuring instrument was improved and its performance wastested. This paper introduced the hardware and software design of the portable SATTmeasuring instrument. The performance of the LED and photosensitive diode weretested. The LED driving circuit and signal amplifying circuit were improved toincrease the light intensity and reduce the measurement error. The prototypeperformance tests indicated that the prototype had good repeatability, stability,consistency and low power consumption.(3) The performance test. We debugged the output of prototype and made sometests about repeatability, stability, consistency, and test power consumption. Theprototype performance tests indicated that the prototype had good repeatability,stability,consistency and low power consumption. Experiments provide a reliableguarantee for the late and lay the foundation for the transition in the future.(4) SATT prediction model was built. Human experiment was designed andconducted on40subjects. The SATTs on20body parts were measured with5different devices.12regression models were established respectively with linear andnonlinear methods. The sub-site model predicted more accurately than the all data model. The upper limb model was the best one, whose correlation coefficient with Bultrasound was0.9. The results showed the sub-site model was more suitable to SATTmeasuring instrument.In summary, the SATT measuring instrument developed in the paper can measureSATT of any body part. It laid the foundation for future production. |