Non-Invasive Blood Glucose Monitor Using Arduino for Clinical Use

Diabetes mellitus is a disease where someone has high blood glucose or low blood glucose. Diabetes mellitus has three categories which are type 1, type 2, and gestational diabetes. Furthermore, people who have diabetes need to check their blood glucose and need to control them frequently. Hence, they need a tool that can be performed by checking the glucose level. However, the conventional method currently used is invasive because blood samples are collected through a finger prick. The method may cause patient discomfort and infection. Thus, this project will establish a non-invasive approach to overcome these concerns caused by invasive methods. The research is focused on people who have type 2 diabetes and fasting state only as it gives better accuracy of blood glucose levels. The research goal is to establish a non-invasive blood glucose monitor to replace a conventional method to detect and display blood glucose measurements via LCD and user phone. The following research objectives are to record glucose level readings in the database system through a cloud network and compare blood glucose measurements with validating results from other journals. The project will use visible red laser light and photodiode to detect the blood glucose level. After that, Arduino converts the equivalent amount of blood glucose sent to the LCD. The glucose level result obtained is transmitted to the cloud network, the Blynk application and ThingSpeak. The project was also able to display the blood glucose level result on the LCD and send the result to the Blynk application and ThingSpeak. The error getting from glucose measurement is not more than 20%, but the average deviation error is 7.08%, where the project can be on par with existing invasive blood glucose measurements. In conclusion, all objectives of this project are achieved.