Model dubokog učenja za klasifikaciju tumora mozga uporabom slika magnetske rezonancije

undergraduate thesis

• Title (english): A deep learning model for brain tumor classification using magnetic resonance images
• Author: Ita Poklepović
• Mentor: Jelena Božek (mentor)
• Committee member: Jelena Božek (predsjednik povjerenstva)
• Committee member: Sonja Grgić (član povjerenstva)
• Committee member: Ivana Žeger (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-11, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj završni rad istražuje primjenu dubokih neuronskih mreža (CNN) na problemu klasifikacije tumora mozga korištenjem slika magnetske rezonancije (MR). Cilj rada bio je upoznati se s postojećim modelima za klasifikaciju tumora na mozgu te implementirati novi model temeljen na stečenom znanju. Detaljno je opisana arhitektura implementiranog modela te njegovi slojevi I razlozi njihovog odabira. Naveden je izvor podataka, njihova podjela i način augmentacije te je opisan proces treniranja s... More unakrsnom validacijom modela. Konačno su analizirani rezultati mreže koristeći navedene metrike te uspoređena točnost modela s javno dostupnim rezultatima drugih modela. Postignuti su odlični rezultati razlikovanja četiri različite vrste tumora s točnošću od čak 99,54\%. Less

Abstract (english)

This bachelor thesis investigates the application of deep neural networks (CNN) to the problem of brain tumor classification using magnetic resonance (MR) images. The aim of the paper was to become familiar with existing brain tumor classification models and to implement a new one based on the acquired knowledge. The architecture of the implemented model, its layers and the reasons for their selection are described in detail. The source and division of data is listed, as well as the... More augmentation method used, and the training and cross-validation process of the model is described. Finally, the results of the network are analyzed using specified metrics and the accuracy of the model was compared with publicly available results of other models. Excellent results were achieved distinguishing four different types of tumors with an accuracy of 99.54\%. Less

• Keywords: duboko učenje; klasifikacija tumora mozga; magnetska rezonanacija; konvolucijska neuronska mreža
• Keywords (english): deep learning; brain tumor classification; magnetic resonance imaging; convolutional neural network
• Language: croatian
• URN:NBN: urn:nbn:hr:168:412014
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2024-10-12 21:53:34