An application that helps diabetics choose their foods by predicting the impact of each food on the sugar before consuming created scientists in the US. The Glucoracle application is based on a customized algorithm and predicting the impact makes eating prognosis easier for diabetics.
Researchers, headed by the biomedical scientist David Albert of the Medical Center of Columbia University in New York, published at the computational biology journal PLoS Computational Biology. Glucoracle allows the patient-user to “upload” blood glucose measurements to the device, as well as pictures of each food and a rough estimate of its nutritional content.
The application then provides the user with an immediate prediction of how much his sugar will increase if he eats that food. To make the first provision, the device should first be “trained” after one week of use by the diabetic, so that the algorithm learns how the body of that person responds to different foods.
The algorithm is continuously updated with new data on how each person’s glucose reacts in each food and thus gradually improves on it’s ability to make predictions. “We have shown that it is possible to create real-time glucose predictions that can be used to people for better dietary choices,” Albers said. Researchers are preparing for a larger clinical trial of the application, which has room for improvement. Clucoracle is expected to be ready for widespread use in about two years.