Aplicación web para determinar el perfil lipídico y su relación con el riesgo cardiovascular en pacientes diabéticos tipo 2 e hipertensos de la consulta externa del Hospital Ceibos
Among the cardiovascular diseases, one of the most common complications in today's society is its high degree of incidence of events with severe and/or lethal outcomes in its convergence population, such as atherosclerosis and dyslipidemia, where the fact of knowing the state of the lipid profi...
Uloženo v:
| Hlavní autor: | |
|---|---|
| Médium: | bachelorThesis |
| Jazyk: | spa |
| Vydáno: |
2022
|
| Témata: | |
| On-line přístup: | https://dspace.unl.edu.ec/jspui/handle/123456789/25246 |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
Přidat tag | Shrnutí: | Among the cardiovascular diseases, one of the most common complications in today's society is its high degree of incidence of events with severe and/or lethal outcomes in its convergence population, such as atherosclerosis and dyslipidemia, where the fact of knowing the state of the lipid profile and cardiovascular risk of patients with diabetes type 2 and systemic arterial hypertension through the information compiled from their records would allow better control of them, focused on a preventive measure that avoids the devaluation of the quality of life. Thus, it is necessary to link computer technology tools with human medicine; therefore, the interaction between these two areas will allow to establish the type of relationship between the lipid profile and cardiovascular risk. On the other hand, this medical-computer study began by manually collecting patient data from the Hospital General Norte de Guayaquil "Ceibos", then performing an automatic cleaning of the same. Later the algorithm of automatic learning of the multiple linear regression was implemented to finish with the grouping of these three previous elements in a web application, where the following results were obtained: of a total of 5617 patient records, 55.59% corresponding to the female gender, and 44.41% to the male gender 2630 are diabetics; and 3052 are hypertensive. Through experimentation, an accuracy of 85.23% of the model was obtained and with the confusion matrix, a 90% effectiveness was obtained, taking a sample of the total set of data collected. Thus, the objective of the degree work was the construction of a website with automatic learning to collaborate together with the internist doctor, providing essential information for decision-making regarding treatments to improve the quality of life of their patients through web development with automatic learning and thereby determine how good or not their lipid profile is associated with their cardiovascular risk. Keywords: Machine Learning (ML), Cholesterol, Cardiovascular Risk. |
|---|