Citas bibligráficas
Trujillo, N., Rojas, E. (2024). Sistema de predicción preliminar para reducir los errores en los diagnósticos de gastroenterología utilizando el algoritmo de Naive Bayes de Machine Learning en un hospital público de Lima [Tesis, Universidad Peruana de Ciencias Aplicadas (UPC)]. http://hdl.handle.net/10757/673568
Trujillo, N., Rojas, E. Sistema de predicción preliminar para reducir los errores en los diagnósticos de gastroenterología utilizando el algoritmo de Naive Bayes de Machine Learning en un hospital público de Lima [Tesis]. PE: Universidad Peruana de Ciencias Aplicadas (UPC); 2024. http://hdl.handle.net/10757/673568
@misc{renati/416983,
title = "Sistema de predicción preliminar para reducir los errores en los diagnósticos de gastroenterología utilizando el algoritmo de Naive Bayes de Machine Learning en un hospital público de Lima",
author = "Rojas Huamán, Eduardo Raúl",
publisher = "Universidad Peruana de Ciencias Aplicadas (UPC)",
year = "2024"
}
Errors in medical diagnoses in gastroenterology have been a recurring problem for patients seeking medical attention. The presence of similar syndromes increases the likelihood of symptom interpretation errors by physicians, who may sometimes not be fully acquainted with certain diseases or medical conditions. Additionally, the lack of communication among healthcare professionals and between physicians and patients, due to congestion in medical care, further exacerbates this situation. These deficiencies can have serious consequences for patients' health, including complications from undiagnosed diseases, lack of prevention of potentially fatal illnesses, and inadequate treatment. Therefore, the aim of this project is to develop a preliminary diagnosis system based on fragmented data to be trained with the Naive Bayes algorithm of Machine Learning, with the primary purpose of assisting specialized physicians in improving the accuracy of their diagnoses. To assess the effectiveness of the proposal, three test cases were conducted, which were evaluated by expert physicians in the field of Gastroenterology. Furthermore, a comparison was made between the proposed system and patients' diagnostic histories, as well as with the Symptomate preliminary diagnostic system from Infermedica, considering three key indicators: user errors, failed prediction attempts, and successful precision attempts. In the first test case, a precision of up to 93% was achieved, demonstrating the system's high reliability in providing preliminary diagnoses, which can be highly beneficial for decision-making by gastroenterology specialists.
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons