Estudio de la evolución de estados prefebriles, para su modelización mediante técnicas de análisis multivariantes
[EN] The research is a process that, by means of the application of the scientific method, procuration obtain information relieving to understand, verify, correct or apply the knowledge. Besides, the research do not belong to an only science, but that can apply in a big numeral of scientific fields....
| Autor: | |
|---|---|
| Tipo de recurso: | tesis doctoral |
| Fecha de publicación: | 2015 |
| País: | España |
| Institución: | Universitat Politècnica de València (UPV) |
| Repositorio: | RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia |
| Idioma: | español |
| OAI Identifier: | oai:riunet.upv.es:10251/51222 |
| Acceso en línea: | https://riunet.upv.es/handle/10251/51222 |
| Access Level: | acceso abierto |
| Palabra clave: | Fiebre Análisis multivariante Análisis discriminante Regresión logística Redes neuronales Predicción ESTADISTICA E INVESTIGACION OPERATIVA |
| Sumario: | [EN] The research is a process that, by means of the application of the scientific method, procuration obtain information relieving to understand, verify, correct or apply the knowledge. Besides, the research do not belong to an only science, but that can apply in a big numeral of scientific fields. In occasions the union of efforts between branches of the science is necessary to resolve some questions, how is the case of this thesis. The collaboration between the medicine and the statistical generates a synergistic effect that allow the obtaining of trustworthy results. With reason of a continuous extension of knowledges in medicine, for can elevate the life average age and the quality of the same, the doctors researchers pose the following hypothesis: the blood cultures of a person, contain a higger quantity of bacteria in a short period of previous time to a feverish beak. How it seems evident, to corroborate this hypothesis would owe compare a series of samples of blood cultures preys in the period pre-febrile and other preys immediately after the feverish beak. The problem that have the doctors researchers is that they do not know when take the samples of blood cultures of the period pre-febrile, since have tools to determine when a person had fever with a sufficient time, and sinus that have the tool to determine when happen a feverish beak in real time (thermometer). In this thesis resolves the problem of when take the sample of blood cultures in the period pre-febrile, by means of mathematical models that feature of measures temperature of the own patient, and a series of variables that are measures of complexity calculated from the same measures of temperature of the patient. The obtaining of the data was done of way non invasive in patients of plant of the hospital of Móstoles, using a system of measure of temperature and storage of data called Thercom c . Once with the available data did a selection of technical of multivariate analysis that could be useful as the type of variables with which works. They calculate different models that can anticipate to a feverish beak, that later validate with new samples of patients. Finally they compare the models on the base of his effectiveness in the prediction of states pre-febrile. The resultant models, are the tool that needed the doctors researchers to take the samples of the period pre-febrile and can continue with his research. |
|---|