Desenvolvimento de Sistema Computacional para Predição da Doença Renal Crônica.

• Ano de defesa: 2020
• Autor(a) principal: MARTINS, Vanessa Edilene Duarte
• Orientador(a): BARROS FILHO, Allan Kardec Duailibe
• Banca de defesa: BARROS FILHO, Allan Kardec Duailibe , SANTANA, Audirene Amorim , SOUSA, Joicy Cortez de Sá , CHAVES, Daniel Praseres , BORGES, Antônio Carlos Romão
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Maranhão
• Programa de Pós-Graduação: PROGRAMA DE PÓS-GRADUAÇÃO EM BIOTECNOLOGIA - RENORBIO/CCBS
• Departamento: DEPARTAMENTO DE ENGENHARIA DA ELETRICIDADE/CCET
• País: Brasil
• Palavras-chave em Português: Aprendizado de máquina; autocuidado; classificadores; diagnóstico precoce; inteligência artificial; medicina computacional
• Palavras-chave em Inglês: Machine learning; self-care; classifiers; early diagnosis; artificial intelligence; computational medicine
• Área do conhecimento CNPq: Modelos Analíticos e de Simulação
• Link de acesso: https://tedebc.ufma.br/jspui/handle/tede/3715

Chronic kidney disease (CKD) does not show signs and / or symptoms in its recent symptoms, and it is important to study and develop alternative methods of diagnosis and / or screening with high sensitivity. Thus, the objective was to develop a computer system for predicting chronic kidney disease. The thesis work is divided into three chapters, in addition to the theoretical foundation and what are the main themes that underlie the present project. Chapter I represents an article entitled “Artificial Intelligence in the Prediction of Chronic Kidney Disease” published in the International Journal of Development Research which aimed to conduct a review of the literature on the use of Artificial Intelligence in the prediction of Chronic Kidney Disease. According to the research, it was observed that CKD can be considered using various classifiers in data mining, as well as predicting the stage of the disease with the use of AI and that the different observed experiences that most classifiers use with high accuracy value, above 90%. Chapter II, a research article entitled “Development of a computer system for tracking patients with chronic kidney disease” published in the International Journal of Development Research aims to build a computer system for the early diagnosis of Chronic Kidney Disease (CKD) using clinical data non invasive, exploring machine learning techniques. And finally, the Chapter III research article “Support System for Chronic Kidney Disease Prediction using Machine Learning” submitted in PeerJ Magazine, aimed to build and validate a chronic kidney disease predictor software related to a classifier algorithm for patients with kidney disease. Among the 3 classifiers used in the experiments, the SVM was the one that obtained the best results and used to test the RDC software predator, demonstrated a good validation performance or can be used in clinical practice as a way of screening patients with disease and for a general population, presenting a low cost and easy execution alternative.