An algorithm based on machine learning has been developed in Novosibirsk to analyze and predict glucose levels in diabetics. This algorithm helps to more accurately predict changes in glucose levels and may be used in the future for precise insulin dosage. This was reported by the press service of the Research Institute of Clinical and Experimental Lymphology (RIKEL), which is a branch of the Institute of Cytology and Genetics of the Siberian Branch of the Russian Academy of Sciences.
The algorithm analyzes continuous glucose monitoring data, identifying glucose dynamics options among all accumulated data. It then determines which dynamics option corresponds to the patient's specific condition and builds a forecast, taking this information into account.
The creation of the algorithm was made possible through the collaboration of scientists from the Research Institute and staff from the Sobolev Institute of Mathematics of the Siberian Branch of the Russian Academy of Sciences. The new approach allows for more accurate prediction of glucose levels, which in the future will open up opportunities for more precise insulin dosage, improving disease management in patients.
In addition, a decision support system for managing glucose levels in patients with type 1 diabetes has been developed. The interactive system asks the doctor for data on the patient's condition and glucose monitoring results, and then provides recommendations for addressing the identified problems.
Read more on the topic:
Russian neural network learns to detect diabetes by human eyes