Synthesis of prostate magnetic resonance images using generative adversarial networks

This master's thesis delves into the realm of medical image synthesis with a particular focus on the application of StyleGAN for generating comprehensive volumetric images of the prostate. The primary objective is to leverage the PI-CAI dataset to create realistic and informative representation...

Descripción completa

Detalles Bibliográficos
Autor: Guardià Olivella, Oriol
Tipo de recurso: tesis de maestría
Fecha de publicación:2024
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/402137
Acceso en línea:https://hdl.handle.net/2117/402137
Access Level:acceso abierto
Palabra clave:Image processing
Imaging systems in medicine
Magnetic resonance
Machine Learning
Deep Learning
GAN
StyleGAN
Medical Imaging
PI-CAI dataset
Imatge -- Generació
Imatges -- Processament
Ressonància magnètica
Imatges mèdiques -- Tractament
Imatgeria mèdica
Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Processament del senyal::Processament de la imatge i del senyal vídeo
Descripción
Sumario:This master's thesis delves into the realm of medical image synthesis with a particular focus on the application of StyleGAN for generating comprehensive volumetric images of the prostate. The primary objective is to leverage the PI-CAI dataset to create realistic and informative representations of prostate anatomy. The research adopts a dual approach: firstly, investigating the utilization of 2D StyleGANs to create reliable prostate MRI 2D slices; secondly, exploring the direct use of 3D models for generating complete volumetric prostate MRI images. The study aims to contribute to the advancement of medical imaging techniques, providing insights into the efficacy of different GAN-based approaches for synthesizing realistic prostate images. The outcomes of this research are expected to be valuable in enhancing medical imaging applications, particularly in the domain of prostate examinations and diagnostics.