Klasifikacija histoloških snimaka crijeva

master's thesis

• Title (english): Classification of Histological Images of the Intestine
• Author: Fran Hrabar
• Mentor: Marko Subašić (mentor)
• Committee member: Marko Subašić (predsjednik povjerenstva)
• Committee member: Zvonko Kostanjčar (član povjerenstva)
• Committee member: Tomislav Petković (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-02-16, Croatia

Abstract

Tema ovog rada je klasifikacija Crohnove bolesti ili ulceroznog kolitisa na temelju histoloških snimki crijeva pomoću dubokih neuronskih mreža.
Rad objašnjava postupak predobrade malog, visokodimenzionalnog i neuravnoteženog skupa podataka kao što su segmentacija i razlamanja snimki na zakrpe radi smanjenja dimenzija ulaza, očuvanja detalja, eliminacije šuma te povećanja primjeraka u skupu podataka te uravnotežavanje skupa podataka i podjelu skupa podataka na skup za treniranje i testiranje. Opisuju se arhitektura modela, proces treniranja i evaluacije modela na temelju odabranih metrika. Metrike pokrivaju razne zahtjeve koje bi liječnik mogao imati pri evaluaciji modela te dopušta stručnjaku da odabere najuspješniji model.

Abstract (english)

The topic of this paper is the classification of Crohn's disease or ulcerative colitis based on histological images of the intestine using deep neural networks.
The paper explains the preprocessing procedure of a small, high-dimensional and imbalanced dataset, such as segmentation and splitting of recordings into patches in order to reduce input dimensions, preserve details, eliminate noise and increase the number of samples in the data set, as well as balancing the data set and dividing the data set into training and testing sets. The architecture of the model, the process of training and evaluation of the model based on selected metrics are described. The metrics cover the various requirements a physician might have when evaluating a model and allow the practitioner to select the most successful model.

• Keywords: umjetna inteligencija; strojno učenje; duboko učenje; računalni vid; umjetne neuronske mreže; duboke neuronske mreže; konvolucijske neuronske mreže; klasifikacija; segmentacija slike; nadzirano učenje; histološke snimke; upalne bolesti crijeva; Crohnova bolest; ulcerozni kolitis; tehnologija u medicini; računalno potpomognuta dijagnostika
• Keywords (english): artificial Intelligence; machine learning; deep learning; computer vision; artificial neural networks; deep neural networks; convolutional neural networks; classification; image segmentation; supervised learning; histological recordings; inflammatory bowel diseases; Crohn's disease; ulcerative colitis; technology in medicine; computer-aided diagnostics
• Language: croatian
• URN:NBN: urn:nbn:hr:168:679913
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2024-03-28 07:46:38