Aplicación de las redes neuronales en el Internet de las Cosas para el diagnóstico de fibrilación auricular
Cargando...
Archivos
Fecha
2022
Autores
Yoliani, E.
Poveda, J.
Martínez, J.
Título de la revista
ISSN de la revista
Título del volumen
Editor
Ediciones Universidad Simón Bolívar
Facultad de Ingenierías
Facultad de Ingenierías
Resumen
El objetivo de este articulo consiste en realizar una revisión y análisis acerca de los distintos métodos de detección de fibrilación auricular, a través de IoT y redes neuronales. En los últimos años, la industria ha generado avances a nivel tecnológico acerca del internet de las cosas y las redes neuronales, traduciéndose en mejores herramientas en distintos campos que suplen la intervención humana y métodos convencionales de los cuales resultan ser costosos y poco prácticos, que en algunos casos no logran detectar la fibrilación auricular correctamente lo cual puede afectar la vida diaria si no es detectada a tiempo. El aprendizaje en línea consiste en entrenar una red neuronal sin disponer previamente de todos los datos. La idea que subyace a este enfoque es que, si no disponemos de todos los datos de entrenamiento, podemos entrenar nuestro sistema para que sea capaz de generalizar lo suficientemente bien para nosotros más adelante, cuando tengamos información más completa sobre lo sucedido durante el historial de cada paciente (como las grabaciones del ECG). Este enfoque ha demostrado ser útil en muchos campos, entre ellos la medicina, donde no siempre hay tiempo para que los pacientes acudan a las clínicas y se sometan a pruebas exhaustivas antes de ser diagnosticados correctamente; por lo tanto, el uso de técnicas de aprendizaje automático podría ayudar a los médicos a diagnosticar afecciones antes de lo que podrían hacerlo de otro modo, al tiempo que se reducen los costes asociados a los procedimientos de pruebas de diagnóstico.
The objective of this article is to review and analyze the different methods of detecting atrial fibrillation using IoT and neural networks. In recent years, the industry has generated technological advances in the internet of things and neural networks, resulting in better tools in different fields that replace human intervention and conventional methods which are expensive and impractical, and in some cases fail to detect atrial fibrillation correctly, which can affect daily life if not detected in time. Online learning consists of training a neural network without first having all the data. The idea behind this approach is that if we do not have all the training data, we can train our system to be able to generalize well enough for us later when we have more complete information about what happened during each patient's history (such as ECG recordings). This approach has proven useful in many fields, including medicine, where there is not always time for patients to come to clinics and undergo extensive testing before being properly diagnosed; therefore, using machine learning techniques could help doctors diagnose conditions earlier than they might otherwise, while reducing the costs associated with diagnostic testing procedures.
The objective of this article is to review and analyze the different methods of detecting atrial fibrillation using IoT and neural networks. In recent years, the industry has generated technological advances in the internet of things and neural networks, resulting in better tools in different fields that replace human intervention and conventional methods which are expensive and impractical, and in some cases fail to detect atrial fibrillation correctly, which can affect daily life if not detected in time. Online learning consists of training a neural network without first having all the data. The idea behind this approach is that if we do not have all the training data, we can train our system to be able to generalize well enough for us later when we have more complete information about what happened during each patient's history (such as ECG recordings). This approach has proven useful in many fields, including medicine, where there is not always time for patients to come to clinics and undergo extensive testing before being properly diagnosed; therefore, using machine learning techniques could help doctors diagnose conditions earlier than they might otherwise, while reducing the costs associated with diagnostic testing procedures.
Descripción
Palabras clave
Fibrilación auricular, Electrocardiograma, Redes neuronales, IoT, Atrial fibrillation, Electrocardiogram, Neural networks
