Publication: Implementación de algoritmos de identificación y predicción para glUCModel
Loading...
Official URL
Full text at PDC
Publication Date
2018
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citations
Exportar
Abstract
La diabetes es una enfermedad que padece un gran porcentaje de la población mundial y que se ha incrementado notablemente en los últimos tiempos por numerosos parámetros como la predisposición genética a sufrir la enfermedad y por otros problemas no genéticos como la alimentación, el estilo de vida sedentario y el aumento de la obesidad.
Las personas que la padecen deben tener un estricto control sobre ella teniendo que hacer un seguimiento diario rutinario de las variaciones de la glucosa en sangre. Aspectos a cuidar con el fin de favorecer glucemias adecuadas serían la alimentación y el ejercicio físico.
En el caso de las personas con diabetes tipo 1 en las que su cuerpo no es capaz de segregar insulina para transformar los carbohidratos en energía, el suministro externo de insulina sintética mediante inyecciones o mediante el uso de una bomba subcutánea de insulina que suministrará de forma periódica dosis en función de la necesidad del cuerpo sería una forma de tener estabilizados estos niveles de glucosa en sangre.
Para poder ayudar a los pacientes a mantener estables sus glucemias se ha incorporado a la aplicación glUCModel, desarrollada por el grupo ABSyS (Adaptive and Bioinspired Systems) del departamento de Arquitectura de Computadores y Automática de la Universidad Complutense de Madrid, un módulo de ayuda a las pautas de insulina que permitirá calcular los niveles de glucosa en sangre en un tiempo determinado máximo no mayor de 3 horas por medio de dos algoritmos pudiendo así dar la posibilidad al usuario de elegir el que más se ajuste a sus necesidades y pueda predecir de forma más exacta.
Los algoritmos utilizados para las predicciones son el del cálculo mediante parámetros terapéuticos, utilizado regularmente por todos los pacientes de diabetes ya que permite calcular de forma rápida la glucosa en sangre después de la ingesta de alimentos teniendo en cuenta la glucosa actual, los carbohidratos que se van a ingerir dando como resultado las dosis de insulina que deben inyectarse para mantener unos niveles óptimos y el algoritmo basado en ARIMA que permite predecir las fluctuaciones de los niveles de glucosa en el organismo de los pacientes únicamente con las variaciones de las glucemias en un tiempo pasado.
Diabetes is a disease that suffers from a large percentage of the world population and which has increased notably in recent times by numerous parameters such as the genetic predisposition to suffer the disease and non-genetic problems such as feeding and sedentary lifestyle and increased obesity. People who have it must have strict control over it by having a daily control blood glucose changes through feeding, physical exercise and in the case of people with type 1 diabetes your body is not able to secrete insulin to transform carbohydrates into energy the external supply of synthetic insulin through injections or by the use of a subcutaneous insulin pump that will periodically deliver doses depending on the need of the body. In order to help patients stay stable their blood sugar has been incorporated into the application glUCModel, developed by the ABSyS group (Adaptive and BioinspiredSystems) of the Department of Computer Architecture and Automation of the Complutense University of Madrid,an insulin aid module that will allow the calculation of blood glucose levels in a maximum time not exceeding 3 hours thanks to two algorithms, thus giving the user the possibility to choose the one that best suits his needs And can predict more accurately. The algorithms used for the predictions are the one of the calculation through therapeutic parameters,used regularly by all diabetes patients since it allows to quickly calculate blood glucose after food intake taking into account the current glucose, the carbohydrates to be ingested resulting in the doses of insulin that must be injected to maintain optimal levels. The algorithm based on ARIMA will also be used to predict the fluctuations of glucose levels in the body of the patients only with the variations of the glycemia in a past time.
Diabetes is a disease that suffers from a large percentage of the world population and which has increased notably in recent times by numerous parameters such as the genetic predisposition to suffer the disease and non-genetic problems such as feeding and sedentary lifestyle and increased obesity. People who have it must have strict control over it by having a daily control blood glucose changes through feeding, physical exercise and in the case of people with type 1 diabetes your body is not able to secrete insulin to transform carbohydrates into energy the external supply of synthetic insulin through injections or by the use of a subcutaneous insulin pump that will periodically deliver doses depending on the need of the body. In order to help patients stay stable their blood sugar has been incorporated into the application glUCModel, developed by the ABSyS group (Adaptive and BioinspiredSystems) of the Department of Computer Architecture and Automation of the Complutense University of Madrid,an insulin aid module that will allow the calculation of blood glucose levels in a maximum time not exceeding 3 hours thanks to two algorithms, thus giving the user the possibility to choose the one that best suits his needs And can predict more accurately. The algorithms used for the predictions are the one of the calculation through therapeutic parameters,used regularly by all diabetes patients since it allows to quickly calculate blood glucose after food intake taking into account the current glucose, the carbohydrates to be ingested resulting in the doses of insulin that must be injected to maintain optimal levels. The algorithm based on ARIMA will also be used to predict the fluctuations of glucose levels in the body of the patients only with the variations of the glycemia in a past time.
Description
Universidad Complutense, Facultad de Informática, curso 2016/2017