Hybrid CNN-GRU Model for Real-Time Blood Glucose Forecasting: Enhancing IoT-Based Diabetes Management with AI.

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Healthcare
• Diabetes Technology

Background:

• Diabetes management requires precise blood glucose level (BGL) monitoring.
• Manual BGL checks are time-consuming and prone to inaccuracies.
• Predicting BGL is complex due to numerous influencing factors.

Purpose of the Study:

• To develop a novel hybrid deep learning model for advanced BGL prediction.
• To integrate this model into Internet of Things (IoT)-enabled diabetes management systems.
• To improve the accuracy and timeliness of BGL forecasting.

Main Methods:

• A hybrid deep learning framework combining Gated Recurrent Units (GRUs) and Convolutional Neural Networks (CNNs) was proposed.
• The GRU layer captures temporal dependencies, while the CNN layer extracts relevant features.
• The model was evaluated using a public type 1 diabetes dataset.

Main Results:

• The hybrid CNN-GRU model demonstrated superior performance compared to standalone LSTM, CNN, and GRU models.
• The proposed model achieved higher prediction accuracy for BGL.
• Real-time data processing on edge devices is facilitated by the model's architecture.

Conclusions:

• The hybrid CNN-GRU model offers a significant advancement in BGL forecasting.
• This approach holds substantial potential for enhancing real-time diabetes management via IoT systems.
• The model's effectiveness suggests improved patient outcomes through more accurate and timely glucose level predictions.