RPI ID: 2013-062-201

Innovation Summary:
This invention involves a predictive algorithm that uses Kalman filtering to analyze glucose readings and forecast hypoglycemic events. The system collects real-time glucose data and estimates current levels and rates of change. By applying predictive models, it calculates future glucose trajectories. If a dangerous drop is anticipated, the system automatically initiates safety measures—such as suspending insulin delivery. This method enhances the performance of insulin pumps by adding a proactive safety layer. The use of Kalman filtering ensures robust performance even with noisy sensor data. This technology represents a major step toward fully closed-loop glucose control systems for diabetic patients.

Challenges / Opportunities:
Many diabetic patients experience unpredictable drops in glucose that current insulin pumps cannot anticipate or prevent. Manual intervention is still required in many cases. This invention solves the need for predictive control in insulin delivery by forecasting dangerous drops in glucose levels. The Kalman filter algorithm is well-suited for processing noisy sensor inputs, making it highly reliable. There is a significant opportunity to enhance patient safety and quality of life through smarter insulin pump systems.

Key Benefits / Advantages:
✔ Predicts glucose crashes before they occur
✔ Automates insulin pump shut-off
✔ Works with noisy or sparse data

Applications:
• Smart insulin pumps
• Continuous glucose monitors (CGMs)
• Closed-loop diabetic therapy systems

Keywords:
#glucosemonitoring #kalmanfilter #insulinpump #diabetes #predictivehealth

Intellectual Property:
US Issued Patent 9,227,014