Distributed Assessment of Virtual Insulin-Pump Settings Using SmartCGMS and DMMS.R for Diabetes Treatment.

Area of Science:

• Biomedical Engineering
• Computational Biology
• Endocrinology

Background:

• Diabetes management relies on insulin pumps for precise glucose control.
• Insulin pump controllers aim for closed-loop systems but require rigorous safety evaluation.
• Current evaluation methods may not cover the full parameter space for safety assessment.

Purpose of the Study:

• To develop and validate a novel computational method for assessing insulin pump controller safety.
• To identify safe and risky combinations of insulin pump parameters using a patient simulator.
• To provide a tool for enhancing the safety and reliability of insulin pump systems.

Main Methods:

• Utilized an FDA-approved diabetic patient simulator on a distributed computing cluster.
• Evaluated a comprehensive Cartesian product of insulin pump parameters.
• Applied binomial models and decision trees to categorize parameter combinations as safe or risky.

Main Results:

• Developed a tool for assessing insulin pump settings and controller safety.
• Demonstrated the tool's efficacy with Low-Glucose Suspend (LGS) and OpenAPS controllers.
• Identified optimal and safe parameter regions, with LGS and OpenAPS showing 1.7 ± 0.6% and 3.2 ± 1.8% local extrema, respectively.

Conclusions:

• The proposed simulation-based method effectively evaluates insulin pump controller safety across a wide parameter range.
• This tool aids in identifying safe insulin pump settings, contributing to improved patient outcomes.
• The methodology provides a scalable approach for the continuous development and validation of diabetes management technologies.