Publication: Sistema avanzado de predicción de crisis migrañosas
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 migraña es una de las enfermedades neurológicas más incapacitantes y está presente en alrededor del 10 % de la población mundial. Sin embargo, aunque se han hecho avances en cuanto a tratamientos de prevención, aún no se ha conseguido crear una cura para esta enfermedad.
Por otra parte, la medicación existente para neutralizar episodios de migraña presenta una característica destacable: es notablemente más efectiva cuando se toma al inicio del episodio. Por el contrario, si se administra cuando ya haya comenzado el dolor su efecto queda reducido (o incluso anulado). No obstante, la mayoría de pacientes de migraña no son capaces de predecir cuando va a comenzar el dolor en un episodio de migraña, por lo que suelen realizar un uso no óptimo de la medicación asociada.
Este proyecto presenta un sistema capaz de predecir la fases de dolor en episodios de migraña. Para ello, se recogen datos de pacientes y se crean modelos predictivos que permiten estimar la probabilidad de aparición de una nueva etapa de dolor. Estas predicciones permiten la generación de alarmas con una antelación de hasta 45 minutos. Por tanto, posibilita un uso adecuado de los medicamentos y aumenta la eficacia de los mismos.
Migraine is one of the most disabling neurological diseases and is present in around 10 % of the world population. However, although progress has been made in terms of prevention treatments, a cure for this disease has not yet been achieved. On the other hand, the existing medication to neutralize migraine episodes presents a remarkable feature: it is quite more effective when taken at the beginning of the episode. On the contrary, if it is administered when the pain has already begun its fe ect is reduced (or even canceled). However, most migraine patients are not able to predict when the pain will begin in a migraine episode, so they usually make a non-optimal use of the associated medication. This project presents a system capable of predicting pain phases in episodes of migraine. For this purpose, patient data is collected and predictive models are created to estimate the probability of appearance of a new pain stage. These predictions allow the generation of alarms up to 45 minutes in advance. Therefore it makes possible an adequate use of the medicines, increasing its effectiveness.
Migraine is one of the most disabling neurological diseases and is present in around 10 % of the world population. However, although progress has been made in terms of prevention treatments, a cure for this disease has not yet been achieved. On the other hand, the existing medication to neutralize migraine episodes presents a remarkable feature: it is quite more effective when taken at the beginning of the episode. On the contrary, if it is administered when the pain has already begun its fe ect is reduced (or even canceled). However, most migraine patients are not able to predict when the pain will begin in a migraine episode, so they usually make a non-optimal use of the associated medication. This project presents a system capable of predicting pain phases in episodes of migraine. For this purpose, patient data is collected and predictive models are created to estimate the probability of appearance of a new pain stage. These predictions allow the generation of alarms up to 45 minutes in advance. Therefore it makes possible an adequate use of the medicines, increasing its effectiveness.
Description
Máster en Ingeniería Informática, Facultad de Informática, Departamento de Arquitectura de Computadores y Automática, curso 2017-2018