Proactive Diabetic Health Monitoring: An Internet of Things–Driven Framework Utilizing a Correlation Attention-Based Dense BiRNN Model

Abstract

Diabetes is a chronic condition demanding continuous management to prevent serious long-term problems. The conventional healthcare approaches depend on irregular clinical visits. It mostly fails to give the constant oversight required for better glycemic control. The Internet of Things (IoT) provides a promising solution by employing connected systems, such as biosensors, to observe crucial medical data in real time. This mechanism supports the secure patient data transmission to medical experts, converting care from reactive to proactive. Likewise, these IoT devices provide patients with immediate feedback, supporting improved self-management. Thus, this research work develops a robust IoT monitoring framework to improve patient outcomes and overall life quality for individuals with diabetes. To initialize the process, the IoT-related data is collected and used by the Correlation Attention-based Dense Bidirectional Recurrent Neural Network (CA-DBiRNN) to monitor the health condition of diabetic patients. This model aids in analyzing the flow of IoT healthcare data more intelligently through the detection of crucial and correlated relationships, thus providing an accurate and rapid intervention for diabetic patients. Finally, the developed model provides the health monitored outcome using the CA-DBiRNN, and the results are compared with some other baseline models, proving its enhanced performance over others.