Detekcija upale pluća na rendgenskim snimkama uporabom metoda dubokog učenja

master's thesis

• Title (english): Pneumonia Detection in Chest X-ray Images Using Deep Learning methods
• Author: Borna Popović
• Mentor: Alan Jović (mentor)
• Committee member: Alan Jović (predsjednik povjerenstva)
• Committee member: Ante Đerek (član povjerenstva)
• Committee member: Željka Mihajlović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-06, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj rad obrađuje detekciju upale pluća na rendgenskim snimkama uporabom metoda dubokog učenja. Upala pluća je respiratorna bolest koju uzrokuju bakterije ili virusi. Analiza rendgenskih snimaka prsnog koša najčešći je način dijagnosticiranja upale pluća. Razvijena je konvolucijska neuronska mreža koja omogućuje automatsku detekciju upale pluća na rendgenskim snimkama prsnog koša. Objašnjena je svrha i funkcionalnost pojedinih slojeva konvolucijskih mreža te su dobiveni rezultati uspoređeni s rezultatima drugih nedavnih radova. Dobivena je točnost od 93.3% na skupu za testiranje. Predložena metoda mogla bi se koristiti samo kao pomoćni alat za donošenje kliničkih odluka. Konačna odluka ipak treba biti prepuštena stručnoj osobi zbog nedostatka jasnog i pravilnog objašnjenja rezultata koje je predvidio model.

Abstract (english)

The subject of this paper is the detection of pneumonia on chest X-rays using deep learning methods. Pneumonia is a respiratory disease caused by a virus or bacteria. Chest X-ray analysis is the most common way to diagnose pneumonia. A convolutional neural network with automatic detection of pneumonia on chest X-rays has been developed. The purpose and functionality of individual CNN layers is explained, and the obtained results are compared with other recent studies. An accuracy of 93.3% was obtained on the test set. The proposed method could only be used as an assistance tool for clinical decision making. However, the final decision should be left to the expert due to the lack of a proper and clear explanation of the results predicted by the model.

• Keywords: upala pluća; rendgenska snimka; duboko učenje; konvolucijska neuronska mreža; implementacija; evaluacija
• Keywords (english): pneumonia; chest X-ray; deep learning; convolutional neural network; implementation; evaluation
• Language: croatian
• URN:NBN: urn:nbn:hr:168:992423
• 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: 2023-01-13 13:32:13