• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.85.2023.tde-12072023-120018
Document
Author
Full name
Caroline Zeppellini dos Santos Emiliozzi
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2023
Supervisor
Committee
Menezes, Mário Olímpio de (President)
Carvalho, Heloisa de Andrade
Fernandes, Leandro Carlos
Title in Portuguese
Aplicação de aprendizado de máquina para melhoria do fluxo de tratamento de radioterapia
Keywords in Portuguese
aprendizado de máquina
ciência de dados
programação linear
radioterapia
tempo de espera
tempo de tratamento
Abstract in Portuguese
O câncer é o principal problema de saúde pública no mundo. A radioterapia é uma das formas mais comuns e efetivas de tratamento de câncer. Porém, atualmente existe um desequilíbrio entre a demanda de tratamentos e a disponibilidade de equipamentos de radioterapia o que leva a atrasos no início de tratamento, esses atrasos produzem sofrimento psicológico e menor probabilidade de controle da doença. Como há uma grande pressão para a contenção de custos, muitas vezes não é possível resolver o problema da falta de equipamentos com a expansão de centros de tratamento. Por outro lado, existe uma ineficiência nos processos relacionados ao fluxo de trabalho e no agendamento de pacientes para início de tratamento. Neste trabalho buscou-se, através da análise de dados do setor de radioterapia do Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo, estudar meios de otimização do fluxo de trabalho para se obter uma gestão eficaz e eficiente do tempo de espera. Com intuito de fazer previsões do tempo de espera e do tempo de tratamento dos pacientes foram comparados quatro algoritmos de AM (Aprendizado de Máquina) com técnica de regressão (Support Vector Machine, Extreme Gradient Boosting , Random forest e Redes neurais) e para a otimização do agendamento de radioterapia foi proposto um modelo de programação linear inteiro misto. Com base no trabalho realizado, conclui-se que a utilização de AM ajuda entender os problemas encontrados no setor. Foram propostas mudanças na rotina, definidos tempo de espera e de tratamento mais adequados e conseguiu-se que o agendamento automático possibilitasse a diminuição do tempo de espera dos pacientes, com priorização dos pacientes com pior prognóstico.
Title in English
Machine learning application to improve the flow of radiotherapy treatment
Keywords in English
data science
linear programming
machine learning
radiotherapy
treatment time
waiting time
Abstract in English
Cancer is the main public health problem in the world. Radiotherapy (RT) is an importante modality in the treatment of these patients. With this growing global burden, demand for RT has been increasing continuously and supply-demand imbalances have become a major concern, due to the negative impact of treatment delays. Evidence has been published of the negative impact of treatment delays on measures such as tumor progression, persistence of cancer symptoms, psychological distress and decreased cancer control and survival rates. The reason for delays in Radiotherapy is not only due to imbalance between capacity and demand, but also due to inefficiency of workflow, for instance, scheduling problems. Consequently, as there is pressure to contain costs, it is often not possible to solve the problem of the lack of equipment. However, the problem of inefficient processes can be attacked. This work, we analyze the data of electronic health records of radiotherapy department of Hospital das Clínicas de São Paulo to attempt to provide a better understanding of the problem and study ways of optimizing workflow and efficient management of waiting time. In order to make predictions of patient waiting time and treatment time, four machine learning algorithms were compared using regression technique (Support Vector Machine, Extreme Gradient Boosting , Random forest and Neural networks ) and for the optimization of radiotherapy scheduling, a mixed integer linear programming model was proposed.Based on this work, it is concluded that the use of AM helps to understand the problems of the departament. Changes in routine were proposed, more appropriate waiting and treatment times were defined and automatic scheduling made it possible to reduce patient waiting times, prioritizing patients with the worst prognosis
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2023-07-12
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.