A joint study by the Institute of Systems Analysis and Informatics of the CNR and theUniversità Campus Bio-Medico di Roma demonstrated how to optimally regulate blood glucose levels in patients with type 1 diabetes using impulse control techniques. The work was published in the IEEE Transactions on Control Systems Technology journal.
Rome, 13 November 2024 - Researchers of theA. Ruberti Institute for Systems Analysis and Computer Science of the National Research Council of Rome (Cnr-Iasi), in collaboration with theUniversity campus Bio-Mbuilding of Rome, have developed a new approach to blood glucose regulation in patients with type 1 diabetes, based on "optimal impulse control" strategies, specifically designed for management via multiple daily injections (MDI). Type 1 diabetes is a chronic disease characterized by persistently high blood glucose levels, caused by a lack of insulin production. Therefore, insulin must be administered externally and on a regular basis, paving the way for the adoption of advanced automatic control techniques. "Optimal impulse control" is an innovative control technique that aims to determine the optimal dose of insulin to inject based on periodic blood glucose measurements, an estimate of blood insulin levels, and the patient's dietary habits.
The method, published in the journal IEEE Transactions on Control Systems Technology, is characterized by the ability to initially calculate an optimal glycemic trajectory offline under ideal conditions. "This calculation is performed through a continuous and nonlinear mathematical model, which represents blood glucose variations as a process influenced by a series of complex and variable factors," explains Alessandro Borri, a researcher at CNR-IASI, explains: "Once the ideal trajectory is defined, the method can track it in real time with minimal computational effort, thanks to an online estimation algorithm that updates itself based on sporadically taken blood glucose measurements." This approach allows the system to adapt and compensate for any non-ideal conditions that arise in daily life, such as eating meals at variable times, with different and unpredictable compositions and caloric intakes.
The control methodology has been validated 'in silico', That is, through computer simulation on a virtual population of one thousand type 1 diabetic patients, generated according to a well-known model approved by the Food and Drug Administration as a substitute for animal testing in preclinical testing of closed-loop blood glucose control strategies. "This preclinical validation campaign confirmed the efficacy of this approach, which promises to improve blood glucose management in patients and optimize the use of insulin therapy, aiming to simplify daily life and improve the well-being of those living with type 1 diabetes," concludes the researcher.
