Klasifikacija slika srčanog tkiva primjenom duboke neuronske mreže

undergraduate thesis

• Title (english): Cardiac Tissue Image Classification Using Deep Neural Network
• Author: Altea Božić
• Mentor: Sven Lončarić (mentor)
• Committee member: Sven Lončarić (predsjednik povjerenstva)
• Committee member: Donik Vršnak (član povjerenstva)
• Committee member: Franko Šikić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-01, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Problemi kliničke dijagnostike, koja se temelji na proučavanju medicinskih slika, često su
klasifikacijske prirode. Jedan takav problem je problem klasifikacije slika srčanog tkiva kod
pacijenata s transplantiranim srcem u svrhu određivanja kliničkog stanja odbacivanja tkiva.
Ovaj rad prati razvoj modela konvolucijske neuronske mreže za dotični problem. Na početku
rada opisane su osnovne značajke konvolucijskih neuronskih mreža. Zatim je objašnjeno
treniranje modela, kao i podatkovni skup korišten za potrebe rada kojeg predstavljaju slike
srčanog tkiva dobivene X-PCI snimanjem. Na kraju, dani su eksperimentalni rezultati
modela uz vizualizaciju uporabom Grad-CAM algoritma te usporedba metrika nekih od
ranije implementiranih modela za dotični problem.

Abstract (english)

The challenges of clinical diagnostics, which rely on the analysis of medical images, are
often classification problems. One such challenge is the classification of cardiac tissue
images in heart transplant patients in order to determine the clinical status of tissue rejection.
This paper outlines the development of a convolutional neural network to address this issue.
The basic characteristics of convolutional neural networks are described at the beginning of
the paper. Then, the training process of the model is explained, along with the description of
the used dataset, which consists of heart tissue images obtained through X-PCI imaging.
Finally, the paper presents experimental results of the model, including visualizations using
the Grad-CAM algorithm and a comparison of metrics with previous versions of the
implemented model for the same problem.

• Keywords: računalni vid; duboko učenje; konvolucijske neuronske mreže; klasifikacija slika; medicinske slike; X-PCI; PyTorch
• Keywords (english): computer vision; deep learning; convolutional neural networks; image classification; medical images; X-PCI; PyTorch
• Language: croatian
• URN:NBN: urn:nbn:hr:168:398639
• 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: Closed access
• Created on: 2024-10-12 20:09:42