prikaz prve stranice dokumenta Generativne suparničke mreže za umjetno generiranje medicinskih slika
  Download
PDF 2.29 MB
master's thesis
Generativne suparničke mreže za umjetno generiranje medicinskih slika
2024. urn:nbn:hr:200:199851
Gavran, Iva
Josip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Department of Software Engineering
Chair of Visual Computing

Cite this document

Gavran, I. (2024). Generativne suparničke mreže za umjetno generiranje medicinskih slika (Master's thesis). Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek. Retrieved from https://urn.nsk.hr/urn:nbn:hr:200:199851

Gavran, Iva. "Generativne suparničke mreže za umjetno generiranje medicinskih slika." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2024. https://urn.nsk.hr/urn:nbn:hr:200:199851

Gavran, Iva. "Generativne suparničke mreže za umjetno generiranje medicinskih slika." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2024. https://urn.nsk.hr/urn:nbn:hr:200:199851

Gavran, I. (2024). 'Generativne suparničke mreže za umjetno generiranje medicinskih slika', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 10 December 2024, https://urn.nsk.hr/urn:nbn:hr:200:199851

Gavran I. Generativne suparničke mreže za umjetno generiranje medicinskih slika [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2024 [cited 2024 December 10] Available at: https://urn.nsk.hr/urn:nbn:hr:200:199851

I. Gavran, "Generativne suparničke mreže za umjetno generiranje medicinskih slika", Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Osijek, 2024. Available at: https://urn.nsk.hr/urn:nbn:hr:200:199851

Please login to the repository to save this object to your list.
Metadata
TitleGenerativne suparničke mreže za umjetno generiranje medicinskih slika
Title (english)Generative adversarial networks for the artificial generation of medical images
AuthorIva Gavran
MentorIrena Galić (mentor)
MentorMarija Habijan (sumentor)
Committee memberMarija Habijan (predsjednik povjerenstva)
Committee memberKrešimir Nenadić (član povjerenstva)
Committee memberKrešimir Romić (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
(Department of Software Engineering)
(Chair of Visual Computing)
Osijek
Defense date and country2024-10-07, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Artificial Intelligence
Abstract
Nedovoljno velik skup podataka jedan je od ključnih izazova kod dubokog učenja, posebno kada je riječ o medicinskim snimcima i pripadajućim im maskama. Zbog ograničenog broja snimaka, postaje vrlo izazovno razviti kvalitetne modele za detekciju rijetkih bolesti. Generiranje umjetnog sadržaja predstavlja jedno od mogućih rješenja za taj problem. U ovom radu obrađena je tematika generiranja novih medicinskih snimaka i pripadajućih im maski pomoću generativne suparničke mreže. Navedeni su i nedostatci tog modela, primjerice zahtjevnost u pogledu vremenskih i računalnih resursa, osjetljivost na parametre te sklonost kreiranju sličnog sadržaja. U završnom dijelu rada predstavljen je model osnovne verzije GAN-a primijenjen za generiranje medicinskih snimaka i maski, prikazani su rezultati generiranja te su istaknuti nedostatci tog modela.
Abstract (english)
Insufficiently large datasets are one of the key challenges in deep learning, especially when it comes to medical images and their corresponding masks. Due to the limited number of images, it is very challenging to develop quality models for detecting rare diseases. Generating artificial content represents one possible solution to this problem. This paper explores the generation of new medical images and their corresponding masks using generative adversarial networks (GANs). The limitations of this model are also discussed, including the high demands in terms of time and computational resources, sensitivity to parameters, and tendency to produce similar content. In the concluding section of the paper, a model based on the basic version of GAN is presented for generating medical images and masks, along with the generated results and the highlighted shortcomings of this model.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:199851
Study programmeTitle: Graduate University Study Programme in Computer Engineering
Study programme type: university
Study level: graduate
Academic / professional title: sveučilišni/a magistar/magistra inženjer/inženjerka računarstva (sveučilišni/a magistar/magistra inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2024-10-30 10:29:04