Non-Invasive IoT-Based Embedded Device for Testing Diabetes Mellitus Type 2 for Early Detection and Prediction

Diabetes is a lifestyle disease caused by poor diet and lifestyle, and if undetected early, it will cause several fatal health issues like heart attacks, strokes, kidney failure, nerve damage, etc. It's a fatal disease that currently has no cure but can be managed if regular testing and proper care are followed. People with diabetes need regular checks of their blood glucose levels to keep track of them and maintain them in healthy ranges. These regular checks are expensive in terms of the consumables being used to perform the tests, i.e., test strips and prick needles, as they are changed every time a test is conducted. Also, the discomfort and pain during the sample collection and the risk of getting an infection are high. This calls for non- invasive methods of blood glucose testing where the test can be done without the need to draw blood. This study uses the Max30102 biosensor, which is a pulse oximeter and heart rate sensor as a non-invasive device to test blood sugar levels by acquiring samples from the patient. The choice of this sensor is informed by studies showing that even during the early stages of diabetes mellitus type 2, there is a change in blood vessels and this in turn affects the pulse rate. The solution showed a positive correlation comparison between the acquired data from volunteers using the non-invasive device and invasive glucometer data from the same volunteers. The results were able to prove the feasibility of the device to test diabetes type 2 without the need to invasively get the samples. However, a lot of data is required to be collected for correlation and comparison using different types of Infrared biosensors having different wavelengths to get high precision and accuracy.