Design and Development of IoT Based Optical Glucometer

Abstract

This research presents the development of an IoT based non-invasive glucometer using optical methods for diabetes monitoring. Diabetes needs to be identified as soon as possible and its development is closely monitored. One of the measures to control this disease is the daily monitoring of blood sugar levels by using glucometer. Glucometers on the market are invasive that require blood sampling or sensors implantation. To get a blood sample, it is necessary to prick a fingertip with a needle to obtain a blood sample. This procedure is uncomfortable, and repeated punctures increase the risk of transmission of infectious diseases. Alternatively, a non-invasive method using optical technique was proposed in this paper. The prototype device is mainly consisting of an NIR LED (940nm) acting as a light transmitter delivered through a finger and reflect to photodetector (BPW34) that acting as a light receiver. The prototype was integrated with the IoT platform using Arduino Cloud for monitoring purpose. The next step involved the development of calibration model. Ten healthy individuals were recruited to take part in glucose readings measurement that conducted by National Kidney Foundation Batu Pahat. A calibration model ( y=82.19x+12.91) was successfully obtained from this experiment. The accuracy of the developed device was between 93.2 and 96.9 % where the error percentage was found to be less than 7 %. In conclusion, a painless non-invasive glucometer based near-infrared LED and photodiode was successfully developed. For future development, the accuracy of the system can possibly be improved by using a longer wavelength of light emitter such as 1500 nm.