Clinical decision support system based on artificial intelligence for adjusting insulin pump parameters in children with type 1 diabetes mellitus

BACKGROUND: Widely available diabetes devices (continuous glucose monitoring, insulin pump etc.) generate large amount of data and development of an advanced clinical decision support system (CDSS), able to automatically evaluate and optimize insulin therapy, is relevant.

AIM: Development of a mathematical model and an CDSS based on it to optimize insulin therapy in children with type 1 diabetes (T1D) and assessment of the agreement between the recommendations of the CDSS and the physician on insulin pump (IP) parameters: basal profile (BP), carbohydrate ratio (CR), correction factor (СF).

MATERIALS AND METHODS: Data from 504 children with T1DM were analyzed over the period of 7875 days. The data included glucose, insulin, food, sex, age, height, weight, diabetes duration and HbA1c. We constructed recurrent neural network (RNN) to predict glucose concentration for 30-120 minutes, an algorithm for optimizing IP settings using prediction results. Next, a software product was developed — a CDSS. To assess the agreement of the recommendations of the CDSS and physicians, retrospective data from 40 remote telemedicine consultations of 40 patients with T1D (median age 11.6 years [7; 15]) were used and 960 points of possible adjustments were analyzed. Three degrees of agreement have been introduced: complete agreement, partial agreement, and complete disagreement. The magnitude of the adjustments was also analyzed.

RESULTS: The accuracy of glycemic predictions was better or comparable with other similar models. The assessment of agreement for BP, CR and CF, according to the Kappa index, showed slight and weak agreement. The frequency of complete agreement between recommendations for adjusting the ongoing IP therapy between the CDSS and physicians is 37.5–53.8%, and complete inconsistency is 4.5–17.4%. From a clinical point of view, consistency in the frequency of occurrence of the indicator is more important. There were no differences in median IP settings between the CDSS and physicians.

CONCLUSION: The CDSS has an acceptable accuracy of glycemic predictions. The CDSS and physicians provide comparable recommendations regarding CSII parameters.

1. «1. American Diabetes Association. Standards of medical care in diabetes 2019. Diabetes Care. 2019;42:1-193

2. Laptev DN, Pereverzeva SV, Emelyanov AO, Peterkova VA. Monitoring of insulin pump therapy in children, adolescents, and young adults with type 1 diabetes mellitus in the Russian Federation. Problems of Endocrinology. 2018;64(2):85- 92. (In Russ.) doi: https://doi.org/10.14341/probl8756

3. Dedov II, Shestakova MV, Peterkova VA, et al. Diabetes mellitus in children and adolescents according to the Federal diabetes registry in the Russian Federation: dynamics of major epidemiological characteristics for 2013–2016. Diabetes mellitus. 2017;20(6):392-402. (In Russ.) doi: https://doi.org/10.14341/DM9460

4. Shestakova MV, Vikulova OK, Zheleznyakova AV, Isakov MA, Dedov II. Diabetes epidemiology in Russia: what has changed over the decade? // Terapevticheskii arkhiv. 2019;91(10):4-13. (In Russ.) doi: https://doi.org/10.26442/00403660.2019.10.000364

5. The DCCT Research Group. Diabetes Control and Complications Trial (DCCT): results of feasibility study. Diabetes Care. 1987;10(1):1-19. doi: https://doi.org/10.2337/diacare.10.1.1

6. The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulindependent diabetes mellitus. N Engl J Med. 1993;329(14):977-986. doi: https://doi.org/10.1056/NEJM199309303291401

7. Pettus J, Edelman SV. Recommendations for using real-time continuous glucose monitoring (rtCGM) data for insulin adjustments in type 1 diabetes. J Diabetes Sci Technol. 2017;11(1):138 -147. doi: https://doi.org/10.1177/1932296816663747

8. Aleppo G, Laffel LM, Ahmann AJ, et al. A practical approach to using trend arrows on the dexcom G5 CGM system for the management of adults with diabetes. J Endocr Soc. 2017;1(12):1445-1460. doi: https://doi.org/10.1210/js.2017-00388

9. Nimri R, Battelino T, Laffel LM, et al. Insulin dose optimization using an automated artificial intelligence-based decision support system in youths with type 1 diabetes. Nature Medicine. 2020;26:1380-1384. doi: https://doi.org/10.1038/s41591-020-1045-7

10. Tyler NS, Mosquera-Lopez CM, Wilson LM, et al. An artificial intelligence decision support system for the management of type 1 diabetes. Nat Metab. 2020;2:612-619. doi: https://doi.org/10.1038/s42255-020-0212-y

11. ISPAD [Internet]. ISPAD Clinical Practice Consensus Guidelines 2022 [cited 2024 May 8]. Available from: https://www.ispad.org/page/ISPADGuidelines2022

12. Romanenkova EM, Eremina IA, Titovich EV, et al. C-peptide levels and the prevalence of islets autoantibodies in children with type 1 diabetes mellitus with different duration of the disease. Diabetes mellitus. 2022;25(2):155-165. (In Russ.) doi: https://doi.org/10.14341/DM12843

13. Laptev DN, Bezlepkina OB, Demina ES, et al. Evaluation of FreeStyle Libre in pediatric t1dm: improved glycemic control, reduction in diabetic ketoacidosis and severe hypoglycemia. Problems of Endocrinology. 2022;68(3):86-92. (In Russ.) doi: https://doi.org/10.14341/probl12877

14. Laptev DN, Emelyanov AO, Andrianova EA, et al. The use of Flash glucose monitoring in children with type 1 diabetes mellitus in real clinical practice. Diabetes mellitus. 2021;24(6):504-510. (In Russ.) doi: https://doi.org/10.14341/DM12817

15. Trufanova ES, Rebrova OY. Decision Support System for Type 1 Diabetes Management. Bachelor’s Thesis. 2021. Federal State Autonomous Educational Institution for Higher Education National Research University Higher School of Economics.

16. Finan DA, Doyle FJ, Palerm CC, et al. Experimental Evaluation of a Recursive Model Identification Technique for Type 1 Diabetes. Journal of Diabetes Science and Technology. 2009;3(5):1192-1202. doi: https://doi.org/10.1177/193229680900300526

17. Fazle R, Yazhou T, Imran H, et al. Stacked LSTM Based Deep Recurrent Neural Network with Kalman Smoothing for Blood Glucose Prediction. BMC Med Inform Decis Mak. 2021;21:101. doi: https://doi.org/10.1186/s12911-021-01462-5

18. Oviedo S, Vehí J, Calm R, Armengol J. A review of personalized blood glucose prediction strategies for T1DM patients. Int J Numer Meth Biomed Engng. 2017;33:2833. doi: https://doi.org/10.1002/cnm.2833

19. Dalla Man С, Rizza R, Cobelli С. Meal simulation model of the glucoseinsulin system. IEEE Trans Biomed Eng. 2007;54(10):1740-1749. doi: https://doi.org/10.1109/TBME.2007.893506

20. Hovorka R, Canonico V, Chassin LJ, et al. Nonlinear model predictive control of glucose concentration in subjects with type 1 diabetes. Physiol Meas. 2004;25(4):905-920. doi: https://doi.org/10.1088/0967-3334/25/4/010

21. Nimri R, Dassau E, Segall T, et al. Adjusting insulin doses in patients with type 1 diabetes who use insulin pump and continuous glucose monitoring: Variations among countries and physicians. Diabetes Obes Metab. 2018;20(10):2458-2466. doi: https://doi.org/10.1111/dom.13408

22. Bachran R, Beyer P, Klinkert C, et al. Basal rates and circadian profiles in continuous subcutaneous insulin infusion (CSII) differ for preschool children, prepubertal children, adolescents and young adults. Pediatr Diabetes. 2012;13(1):1-5. doi: https://doi.org/10.1111/j.1399-5448.2011.00777.x

23. Heinemann L, Nosek L, Kapitza C, et al. Changes in basal insulin infusion rates with subcutaneous insulin infusion: time until a change in metabolic effect is induced in patients with type 1 diabetes. Diabetes Care. 2009;32(8):1437-1439. doi: https://doi.org/10.2337/dc09-0595

24. Danne T, Battelino T, Jarosz-Chobot P, et al. Establishing glycaemic control with continuous subcutaneous insulin infusion in children and adolescents with type 1 diabetes: experience of the PedPump Study in 17 countries. Diabetologia. 2008;51(9):1594-1601. doi: https://doi.org/10.1007/s00125-008-1072-2

25. Nimri R, Muller I, Atlas E, et al. MD-Logic Overnight Control for 6 Weeks of Home Use in Patients With Type 1 Diabetes: Randomized Crossover Trial. Diabetes Care. 2014;37(11):3025-3032. doi: https://doi.org/10.2337/dc14-0835

26. Simon AC, Schopman JE, Hoekstra JB, et al. Factors that drive insulin-dosing decisions of diabetes care providers: a vignettebased study in the Netherlands. Diabet Med. 2015;32(1):69-77. doi: https://doi.org/10.1111/dme.12586