Please use this identifier to cite or link to this item:
Title: Cálculo de dimensiones de regiones de interés del cerebro en resonancias magnéticas
Author: Segura Anaya, Miguel
Director: Ventura Royo, Carles  
Tutor: Kanaan Izquierdo, Samir
Others: Universitat Oberta de Catalunya
Keywords: machine learning
convolutional neural network
image diagnosis
Issue Date: 5-Jun-2018
Publisher: Universitat Oberta de Catalunya
Abstract: The use of information technology as a tool to help medicine is a reality. The management of information is something integrated in clinical practice. Some of the most well-known applications are found in diagnostic imaging. Due to the quality of the image and sensitivity, MRI is currently considered the best test to evaluate abnormalities or disorders in humans. One of the latest trends in the field of bioinformatics is the application of "machine learning" as a tool for evaluating these medical tests. This work focuses on exploring this field and defining a framework where we can merge these two technologies. Taking as initial reference FreeSurfer, one of the most popular tools in the study of MRI, we will try to obtain the same results using an artificial intelligence system. We will start our work analyzing the capabilities of FreeSurfer and selecting among all the measures that calculate our initial objective. Next, we will analyze the options of data sources and the format of these data in order to have enough information to be able to train and verify our systems. Finally, we will select a work environment for our artificial intelligence models, we will configure them and train them in order to discover the level of precision that our results may have. The work ends with an evaluation of the obtained results, a list of conclusions and proposals to evolve the system.
Language: Spanish
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
seganamiTFM0618memoria.pdfMemoria del TFM1.39 MBAdobe PDFView/Open

This item is licensed under a Creative Commons License Creative Commons