A quick diagnosis of the type 1 diabetes mellitus (DM1) through non-invasive means is an ideal way to provide a detection mechanism for paramedics before the involvement of GPs or other specialty staffs. Primarily the diabetic test has been obtained with the urinary and blood ketone sample, however these methods are invasive, expensive, time consuming and inconvenient. The chemical decomposition analysis of organic compounds in the breath can provide an indication of glucose level. With the help of a suitable VOC sensor connected to Arduino and some data processing, the composition of ketone bodies, ethanol, methanol and other exogenous compounds can be analysed. The monitoring of the ketone bodies provides an efficient means of monitoring diabetic level for itsconvenience and non-invasive diagnosis. Here the VOC data after its acquisition and processing using Arduino Uno has been transmitted to a mobile phone via Bluetooth and NFC communication and the result has been analysed and displayed.We plan to run some machine learning algorithm over android development as part the data processing and analysis. The diabetic level would be compared with data obtained from other invasive methods.
Published July 29, 2019