Meal-Related Glycemic Trend Information to Assist Bolus Decision-Making in People with Type 1 Diabetes / Carlos Esteban Builes Montaño ; Lema Pérez, L. ; Álvarez, L. ; García Tirado, J.  

Público ISBD Título : Meal-Related Glycemic Trend Information to Assist Bolus Decision-Making in People with Type 1 Diabetes Tipo de documento : documento electrónico Autores : Carlos Esteban Builes Montaño, Autor ; Lema Pérez, L., Autor ; Álvarez, L., Autor ; García Tirado, J., Autor Fecha de publicación : 2024 Títulos uniformes : ScienceDirect Idioma : Inglés (eng) Palabras clave : Diabetes mellitus; Computer Simulation; Mathematical model; Simulation; Decision support systems; Training Resumen : Carbohydrate counting is an essential skill of type 1 diabetes therapy. However, this method is based on individual estimates and is prone to errors. Moderate-to-large estimate errors may lead to postprandial hyperglycemia (carbohydrate underestimation) or hypoglycemia (carbohydrate overestimation), both undesired therapy side effects. In this manuscript, we present an analysis of the prediction ability of a decision support system that uses a mathematical model representing the main organs involved in glucose metabolism to simulate the glycemic trends of meals about to be eaten. This decision support system aims to reinforce mindful eating and sharpen the user's informed decision-making process regarding meal bolus. We use data from 5 participants (continuous glucose monitoring, food intake, and insulin records) from a previous study for comparison purposes (ClinicalTrials.govNCT03859401). Mean absolute error and mean absolute relative difference were computed to assess the prediction ability of the proposed method. Limits of agreement for the readings were obtained using a non-parametric approach. Finally, a clinical assessment was conducted using an error grid analysis. Compared to the CGM data, the mean absolute error was 15.32 ± 23.02 mg/dL with a mean absolute relative difference of 10.12%. 95% of the differences were inferior to 50 mg/dL. The accuracy of the model as per the grid analysis was high, with more than 99% of the readings falling in the safe zones (A or B). The presented method paves a new way to improve carbohydrate counting and help people with diabetes manage meal uncertainty. The method will be tested in a subsequent clinical study. Mención de responsabilidad : C. Builes Montaño, L. Lema Perez, H. Alvarez, J. Garcia Tirado Referencia : IFAC-PapersOnLine Volume 58, Issue 24, 2024, Pages 321-326 DOI (Digital Object Identifier) : 10.1016/j.ifacol.2024.11.057 Derechos de uso : CC BY-NC-ND En línea : https://www.sciencedirect.com/science/article/pii/S2405896324021840 Enlace permanente : https://hospitalpablotobon.cloudbiteca.com/pmb/opac_css/index.php?lvl=notice_dis Meal-Related Glycemic Trend Information to Assist Bolus Decision-Making in People with Type 1 Diabetes [documento electrónico] / Carlos Esteban Builes Montaño, Autor ; Lema Pérez, L., Autor ; Álvarez, L., Autor ; García Tirado, J., Autor . - 2024. Obra : ScienceDirect Idioma : Inglés (eng) Palabras clave : Diabetes mellitus; Computer Simulation; Mathematical model; Simulation; Decision support systems; Training Resumen : Carbohydrate counting is an essential skill of type 1 diabetes therapy. However, this method is based on individual estimates and is prone to errors. Moderate-to-large estimate errors may lead to postprandial hyperglycemia (carbohydrate underestimation) or hypoglycemia (carbohydrate overestimation), both undesired therapy side effects. In this manuscript, we present an analysis of the prediction ability of a decision support system that uses a mathematical model representing the main organs involved in glucose metabolism to simulate the glycemic trends of meals about to be eaten. This decision support system aims to reinforce mindful eating and sharpen the user's informed decision-making process regarding meal bolus. We use data from 5 participants (continuous glucose monitoring, food intake, and insulin records) from a previous study for comparison purposes (ClinicalTrials.govNCT03859401). Mean absolute error and mean absolute relative difference were computed to assess the prediction ability of the proposed method. Limits of agreement for the readings were obtained using a non-parametric approach. Finally, a clinical assessment was conducted using an error grid analysis. Compared to the CGM data, the mean absolute error was 15.32 ± 23.02 mg/dL with a mean absolute relative difference of 10.12%. 95% of the differences were inferior to 50 mg/dL. The accuracy of the model as per the grid analysis was high, with more than 99% of the readings falling in the safe zones (A or B). The presented method paves a new way to improve carbohydrate counting and help people with diabetes manage meal uncertainty. The method will be tested in a subsequent clinical study. Mención de responsabilidad : C. Builes Montaño, L. Lema Perez, H. Alvarez, J. Garcia Tirado Referencia : IFAC-PapersOnLine Volume 58, Issue 24, 2024, Pages 321-326 DOI (Digital Object Identifier) : 10.1016/j.ifacol.2024.11.057 Derechos de uso : CC BY-NC-ND En línea : https://www.sciencedirect.com/science/article/pii/S2405896324021840 Enlace permanente : https://hospitalpablotobon.cloudbiteca.com/pmb/opac_css/index.php?lvl=notice_dis

Reserva

Reservar este documento

Ejemplares(1)

Código de barras	Número de Ubicación	Tipo de medio	Ubicación	Sección	Estado
DD002310	AC-2024-165	Archivo digital	Producción Científica	Artículos científicos	Disponible

---

1 (1 - 1 / 1) Por página: 25 50 100 200