Dia-Tracker: a Multimodal Diabetes Assistance Device

Diabetes continues to emerge as perilous global health issue. It is anticipated that $\mathbf{5 5 2}$ million adults will receive a diabetes diagnosis by 2030. The principal cause is unregulated glucose levels. The infrared and red LED-based pulse oximeter achieved significant global appeal during the COVID-19 pandemic. Many investigations have aimed to incorporate noninvasive glucose monitoring with oxygen saturation (SpO2), heart rate (HR) in pulse oximeters. We aim to create a prototype model capable of simultaneously measuring heart rate, SpO2, glucose levels, and cholesterol levels while incorporating an insulin infusion system for diabetes monitoring. Data were collected from eight individuals to ascertain the regression equation utilizing infrared and red response values. The system's accuracy is confirmed by comparing HR & SpO2 with commercial pulse oximeter. The sensor demonstrates a 95% accuracy in functionality, as determined by heart rate and $\text{SpO}_{2}$ measurements. The $\mathrm{R}^{2}$ values for predicting blood glucose and cholesterol from IR and RED LED responses are 0.7394 and 0.9525, respectively. User comfort is evaluated highly, and the insulin infusion device attains $\mathbf{9 0 \%}$ accuracy.