Loading...
Publication
Técnicas avançadas de inteligência artificial para a deteção e rastreio de doenças gastrointestinais
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 7.34 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
As doenças gastrointestinais têm vindo a aumentar devido a vários fatores associados ao estilo
de vida moderno, como por exemplo uma alimentação inadequada, sedentarismo e tabagismo.
A colonoscopia continua a ser o método de referência para o diagnóstico de patologias
intestinais, permitindo a deteção e tratamento de lesões. No entanto, a sua precisão depende
fortemente da experiência do médico, resultando em variabilidade nos diagnósticos e
potenciais atrasos na deteção de condições críticas. Para além disso, o procedimento de
colonoscopia pode exigir a realização de biópsias invasivas, que, embora essenciais para
diagnóstico definitivo, acarretam riscos e desconforto para os pacientes.
Esta dissertação de tese explora a integração de técnicas de visão computacional e deep
learning para otimizar a análise de colonoscopias, com o objetivo de melhorar a precisão na
deteção de lesões e apoiar a tomada de decisão clínica. Através do uso de redes neuronais
convolucionais (CNNs) e modelos de segmentação como o ResNet, DenseNet e Inception, esta
investigação propõe o desenvolvimento de um sistema baseado em inteligência artificial capaz
de identificar e classificar lesões colorretais com maior precisão e consistência. O sistema
proposto visa complementar a experiência médica, reduzindo a variabilidade nos diagnósticos
e otimizando os processos de rastreio.
Os resultados desta dissertação mostram que os modelos híbridos, que combinam diferentes
arquiteturas convolucionais, superaram os modelos baseados apenas em transfer learning, que
apresentaram desempenhos insatisfatórios. A melhor performance foi alcançada pelo modelo
híbrido ResNet + EfficientNet + DenseNet, com accuracy de 86,67%. Esses resultados sugerem
que a abordagem híbrida é mais eficaz para a deteção de lesões gastrointestinais, podendo
contribuir para diagnósticos mais rápidos e precisos, além de reduzir a necessidade de biópsias
desnecessárias.
Gastrointestinal diseases have been increasing due to various factors associated with modern lifestyle, such as inadequate diet, sedentary behavior, and smoking. Colonoscopy remains the gold standard method for diagnosing intestinal pathologies, allowing for the detection and treatment of lesions. However, its accuracy heavily depends on the physician’s experience, leading to variability in diagnoses and potential delays in detecting critical conditions. Additionally, the colonoscopy procedure may require the performance of invasive biopsies, which, although essential for definitive diagnosis, carry risks and discomfort for the patients. This thesis explores the integration of computer vision and deep learning techniques to enhance colonoscopy analysis, aiming to improve lesion detection accuracy and support clinical decisionmaking. By leveraging convolutional neural networks (CNNs) and segmentation models like UNet, this research seeks to develop an AI-based system capable of identifying and classifying colorectal lesions with higher precision and consistency. The proposed system aims to complement medical expertise, reducing diagnostic variability and optimizing screening processes. The results of this dissertation show that hybrid models, which combine different convolutional architectures, outperformed models based solely on transfer learning, which showed unsatisfactory performance. The best performance was achieved by the hybrid model ResNet + EfficientNet + DenseNet, with an accuracy of 86.67%. These results suggest that the hybrid approach is more effective for the detection of gastrointestinal lesions, potentially contributing to faster and more accurate diagnoses, while also reducing the need for unnecessary biopsies.
Gastrointestinal diseases have been increasing due to various factors associated with modern lifestyle, such as inadequate diet, sedentary behavior, and smoking. Colonoscopy remains the gold standard method for diagnosing intestinal pathologies, allowing for the detection and treatment of lesions. However, its accuracy heavily depends on the physician’s experience, leading to variability in diagnoses and potential delays in detecting critical conditions. Additionally, the colonoscopy procedure may require the performance of invasive biopsies, which, although essential for definitive diagnosis, carry risks and discomfort for the patients. This thesis explores the integration of computer vision and deep learning techniques to enhance colonoscopy analysis, aiming to improve lesion detection accuracy and support clinical decisionmaking. By leveraging convolutional neural networks (CNNs) and segmentation models like UNet, this research seeks to develop an AI-based system capable of identifying and classifying colorectal lesions with higher precision and consistency. The proposed system aims to complement medical expertise, reducing diagnostic variability and optimizing screening processes. The results of this dissertation show that hybrid models, which combine different convolutional architectures, outperformed models based solely on transfer learning, which showed unsatisfactory performance. The best performance was achieved by the hybrid model ResNet + EfficientNet + DenseNet, with an accuracy of 86.67%. These results suggest that the hybrid approach is more effective for the detection of gastrointestinal lesions, potentially contributing to faster and more accurate diagnoses, while also reducing the need for unnecessary biopsies.
Description
Keywords
Visão Computacional Deep Learning Tranfer Learning Explainable IA Segmentação de Imagens Lesões colorretais Colonoscopias Visão computacional Segmentação de imagens Lesões colorretais Colonoscopias