BACKGROUND: Closed-loop control algorithms in diabetes aim to calculate the optimum insulin delivery to maintain the patient in a normoglycemic state, taking the blood glucose level as the algorithm's main input. The major difficulties facing these algorithms when applied subcutaneously are insulin absorption time and delays in measurement of subcutaneous glucose with respect to the blood concentration. METHODS: This article presents an inverse controller (IC) obtained by inversion of an existing mathematical model and validated with synthetic patients simulated with a different model and is compared with a proportional-integral-derivative controller. RESULTS: Simulated results are presented for a mean patient and for a population of six simulated patients. The IC performance is analyzed for both full closed-loop and semiclosed-loop control. The IC is tested when initialized with the heuristic optimal gain, and it is compared with the performance when the initial gain is deviated from the optimal one (+/-10%). CONCLUSIONS: The simulation results show the viability of using an IC for closed-loop diabetes control. The IC is able to achieve normoglycemia over long periods of time when the optimal gain is used (63% for the full closed-loop control, and it is increased to 96% for the semiclosed-loop control).
BACKGROUND: Closed-loop control algorithms in diabetes aim to calculate the optimum insulin delivery to maintain the patient in a normoglycemic state, taking the blood glucose level as the algorithm's main input. The major difficulties facing these algorithms when applied subcutaneously are insulin absorption time and delays in measurement of subcutaneous glucose with respect to the blood concentration. METHODS: This article presents an inverse controller (IC) obtained by inversion of an existing mathematical model and validated with synthetic patients simulated with a different model and is compared with a proportional-integral-derivative controller. RESULTS: Simulated results are presented for a mean patient and for a population of six simulated patients. The IC performance is analyzed for both full closed-loop and semiclosed-loop control. The IC is tested when initialized with the heuristic optimal gain, and it is compared with the performance when the initial gain is deviated from the optimal one (+/-10%). CONCLUSIONS: The simulation results show the viability of using an IC for closed-loop diabetes control. The IC is able to achieve normoglycemia over long periods of time when the optimal gain is used (63% for the full closed-loop control, and it is increased to 96% for the semiclosed-loop control).
Authors: Alexis Alonso-Bastida; Manuel Adam-Medina; Rubén Posada-Gómez; Dolores Azucena Salazar-Piña; Gloria-Lilia Osorio-Gordillo; Luis Gerardo Vela-Valdés Journal: Int J Environ Res Public Health Date: 2022-01-09 Impact factor: 3.390