Descubrimiento de conocimiento en entornos hospitalarios a partir de registros médicos para la toma de decisiones
Fecha
2016-05-25
Autores
Fuentes Herrera, Ivett Elena
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Central “Marta Abreu” de Las Villas
Resumen
Debido al incremento exponencial de la información almacenada en las organizaciones, la Sociedad de la Información está siendo superada por la necesidad de nuevos métodos capaces de procesar la información y asegurar su uso productivo. Esto se hace lógicamente extensible a los centros hospitalarios, a partir del uso extendido de las Historias Clínicas en formato electrónico. Disponer de información sistematizada, gestionarla de forma eficiente y segura es esencial para garantizar mejores prácticas en salud. A esto se le añade, la necesidad de soportar estándares que permitan el intercambio entre las instituciones de salud; específicamente HL7 es de los más utilizados debido a que proporciona el intercambio a partir del meta-lenguaje XML.
En este trabajo se propone un modelo para el descubrimiento de conocimiento implícito en Historias Clínicas Electrónicas en formato semi-estructurado. Se define una variante para asistir la toma de decisiones diagnóstica de un nuevo paciente, mediante una clasificación supervisada que utiliza Razonamiento Basado en Casos para la clasificación de documentos. Se desarrolló el paquete denominado CDADM1, que implementa el modelo presentado. Este sistema, permite la manipulación de Historias Clínicas a partir del estándar CDA, integrando agrupamiento y clasificación automática en dependencia de los criterios del decisor y con ello, asiste al personal médico en la toma de decisiones y facilita el entrenamiento asistencial de estudiantes en los centros hospitalarios.
Due to the exponential increase in the organizations of stored information, the Information Society is being overtaken by the need for new methods capable of processing information and ensuring productive use. This is logically extended to the hospitals, from the widespread use of clinical histories in electronic format. To have systematized information, manage it efficiently and securely is essential to ensure better health practices. In addition, there is the need for standards to support the exchange among health institutions; specifically HL7 has become one of the most widely used because it provides the exchange from XML. This study presents a methodology for the discovery of implicit knowledge Medical Records in semi-structured format using the content and structure of these is proposed. A variant is defined to assist in making diagnostic decisions a new patient by a supervised classification using RBC for the recovery and classification of documents; and the conversion of textual representation of cases used to retrieve documents similar documents. A package called CDADM was implemented, which the methodology presented. The system integrates the methodology and the automatic classification allows the manipulation of medical records from the CDA standard, hence contributing to information and knowledge management in health institutions.
Due to the exponential increase in the organizations of stored information, the Information Society is being overtaken by the need for new methods capable of processing information and ensuring productive use. This is logically extended to the hospitals, from the widespread use of clinical histories in electronic format. To have systematized information, manage it efficiently and securely is essential to ensure better health practices. In addition, there is the need for standards to support the exchange among health institutions; specifically HL7 has become one of the most widely used because it provides the exchange from XML. This study presents a methodology for the discovery of implicit knowledge Medical Records in semi-structured format using the content and structure of these is proposed. A variant is defined to assist in making diagnostic decisions a new patient by a supervised classification using RBC for the recovery and classification of documents; and the conversion of textual representation of cases used to retrieve documents similar documents. A package called CDADM was implemented, which the methodology presented. The system integrates the methodology and the automatic classification allows the manipulation of medical records from the CDA standard, hence contributing to information and knowledge management in health institutions.
Descripción
Palabras clave
Descubrimiento de Conocimiento, Registros Médicos, Historias Clínicas Electrónicas, Razonamiento Basado en Casos, Clasificación Automática, Toma de Decisiones, Aprendizaje Automatizado, Sistemas de Información Hospitalarios, Inteligencia Artificial