Smart Artificial-Intelligence Based Self-Care-Device for Diabetic Patients

Abstract

The prevalence of diabetes has been consistently increasing over the past decades. While implementation of sensor-based systems for managing diabetes is on the rise, manual monitoring efforts are still required in case of abnormal fluctuations in the blood glucose levels due to lack of intelligent systems. This paper presents an incremental research work developed on top of a previous work to build a smart system. The proposed system helps in managing Type-1 diabetes using future predictions to avoid hypoglycemia or hyperglycemia. It presents a smart healthcare device for diabetic patients with the ability of intelligently controlling the insulin dosage each time during automatic injection to maintain normal blood glucose levels. The system consists of an insulin pump, a continuous glucose monitor (CGM), and a microprocessor unit which utilizes artificial intelligence (AI) based algorithms to predict future blood glucose levels to avoid serious conditions. The prediction feature used in the proposed system distinguishes it from other commercially available products. Furthermore, the proposed system monitors the blood glucose levels of the patient and notifies the patient in case of critical conditions. It can also send a summarized report to the healthcare center and provide the patient with a timeline of collected data and insulin dosages injected. The device was implemented and tested with very good results.