Sistema automatizado para pronósticos en pacientes con problemas renales
Fecha
2007-06-28
Autores
Leyva Cabrera, Haydee
Martín Pérez, Maybel
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central “Marta Abreu” de Las Villas
Resumen
El problema de pronosticar la esperanza de vida de un paciente en el momento de su ingreso a la sala de Nefrología es un dato de interés, tanto del personal médico como del paciente pues es un medidor de la eficacia del tratamiento que se brinda a los pacientes con insuficiencia renal.
Utilizamos para este propósito como técnica de inteligencia artificial los sistemas basados en casos pues se ajusta de manera natural al procedimiento de un médico al enfrentarse a la situación de clasificar la esperanza de vida de un paciente en determinados rangos. Se estudió la posibilidad de implementar en Weka una función de distancia con pesaje de rasgos que podría elevar los resultados de dicha clasificación.
Por otra parte la decisión de si un paciente es transplantable o no requiere por parte de los expertos un análisis de varios factores por lo cual se hace engorroso. Para brindarles un apoyo a la toma de decisiones implementamos un clasificador simple basado en reglas que ofrece un resultado atendiendo a las especificaciones establecidas y que además en el caso de no ser apto para transplante brinda un reporte sencillo de las causas por las que no debe ser sometido a dicho tratamiento.
The fact of predict the life expectancy of a patient when entered to the nephrology’s room is undoubtedly an interesting data for the doctors and for the patients as well, due its characteristics of bringing some measure of the efficacy of the treatment for people with Chronic Renal Failure . We used for achieve this goal as a Artificial Intelligence technique the case-based reasoning, because it fits in a natural way to the doctor's procedure to classifying life expectancy of a patient in given ranges. We studied the possibility of a further implementation in Weka of a distance function with features weighting that could improves the results of such classification. On the other hand the decision of if a patient is transplantable require for the expert’s part the analysis of several factors whereby it is made bothersome, to offer a decision makings support we implemented a simple rules based classifier that offers a result attending to the established specifications and that besides in the event of not being apt it offers a simple report of the causes for which he must not receive this treatment.
The fact of predict the life expectancy of a patient when entered to the nephrology’s room is undoubtedly an interesting data for the doctors and for the patients as well, due its characteristics of bringing some measure of the efficacy of the treatment for people with Chronic Renal Failure . We used for achieve this goal as a Artificial Intelligence technique the case-based reasoning, because it fits in a natural way to the doctor's procedure to classifying life expectancy of a patient in given ranges. We studied the possibility of a further implementation in Weka of a distance function with features weighting that could improves the results of such classification. On the other hand the decision of if a patient is transplantable require for the expert’s part the analysis of several factors whereby it is made bothersome, to offer a decision makings support we implemented a simple rules based classifier that offers a result attending to the established specifications and that besides in the event of not being apt it offers a simple report of the causes for which he must not receive this treatment.
Descripción
Palabras clave
Esperanza de Vida, Pronóstico, Pacientes con Problemas Renales, Razonamiento Basado en Casos, Sistema Automatizado, Inteligencia Artificial