Preliminary Study Of The Exhaled Gas Diagnosis Of Diabetes

Diabetes is a chronic disease, a common non-infectious serious hazards to humanhealth and modern medicine has no cure diabetes, blood glucose levels in patientswithfrequent determination is an important means to delay diabetes and itscomplications.This paper summarizes the exhaled air non-contact detectionof disease, especially the application of the detection of diabetes. Based on thecolorimetric array sensors for gas recognition mechanism of response, and combinedwith electronic and detection methods, and design diabetes exhaled gas detectionequipment.Firstly, the gas is pumped into the reaction sealing device. Secondly, the images ofcolorimetric sensor array before and after responses to the gas are obtained by theoptical sensor. And then, the color changes of the colorimetric sensor array are acquiredthrough a series of image processing. To be the end, the detection results are obtained.The detailed works of this paper include:①Based on the requirements of the diabetes breath detection equiment, modulardesign is applied in the structure of the hardware. As a host controller, embeddedARM9system controls all the working process; as service controller, PIC16F877Acontrols the external modules, which include Micro-pump driver module,monitoring modules (eg. temperature, flow, and humidity), the LED driver moduleand the USB communication module.②According to theprincipleand process of the gasdetection, and detection systemsoftware analysis, each module of the software was design.③Acetone gas was preparation and deployed into a gradient of concentration,provided for the detection.④The gas was researched with the developed exhaled gas detection device based oncolorimetric sensor array. The response time acetone gas of is less than30s. Usingcluster analysis and discriminant analysis, the accuracy of identification fordifferent concentrations was100%. It was also shown that the gas with differentconcentrationcan also be reliably classified with100%accuracy by SVM. Usingfuzzy neural network for quantitative identification, the results shown thatthe error withinthe allowable.This study showed that developed exhaled gas detection device based oncolorimetric sensor array can be used for the rapid, sensitive, and precise detection of acetone gas. Comparing whith large analytic apparatus, the colorimetric sensor arrayhas the advantages of low-cost and easy operating, and it may become an importantdetecting method, and provide ideas and directions for further in-depth study...