Algorithm helps make continuous glucose monitoring more accurate

Two universities in Spain have patented a new system to more accurately monitor blood glucose in Type 1 diabetes. The work, from the Universitat Politècnica de València and the Universitat de Girona, is published at Biomedical and Health Informatics.

It's based on a new calibration algorithm adapted to existing devices for controlled and automatic insulin release, according to an announcement posted to MedicalXpress.

Intensive insulin therapy requires multiple daily injections or continuous infusion with an insulin pump. However, it increases the episodes of hypoglycemia, which carries serious potential side effects including diabetic coma.

València researcher Jorge Bondia calls the method a step toward an artificial pancreas for automatic glycemic control. The new algorithm aims to reduce errors in estimated glucose levels in continuous monitoring based on data from a representative population of patients.

"Current algorithms are based on linear regression techniques where the dynamic information between different biological compartments is ignored and this may cause high estimation errors. It is precisely the magnitude of the error that caused the continuous glucose monitoring to be considered today as a tool to complement and not replace capillary measurement," Bondia said, and he wants to change that.

Continuous glucose monitoring devices and insulin pumps used together are more effective at managing Type 1 diabetes and they provide patients a better quality of life, Johns Hopkins University researchers found. Though they're more expensive, they found patients prefer to use the technology.

Remote diabetes monitoring is expected to become the second-most-common use of telehealth--after monitoring of congestive heart failure patients--by 2017, according to market research firm InMedica. In all, it estimates 1.8 million patients will be treated through telehealth in the next five years.

Other research has focused on ways to eliminate needles from glucose monitoring, including the use of biosensors that can take a reading from sweat or tears.

To learn more:
- find the research
- here's announcement