Usporedba modela strojnog učenja za klasifikaciju poremećaja iz spektra autizma korištenjem slika funkcijske magnetske rezonancije mozga

undergraduate thesis

• Title (english): Comparison of machine learning models for the autism spectrum disorder classification using brain functional magnetic resonance images
• Author: Dominik Pavelić
• 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: 2024-07-05, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Poremećaji iz spektra autizma jedni su od najzastupljenijih neurorazvojnih poremećaja, tradicionalne metode dijagnoze često su spore i subjektivne. U ovome radu uspoređene su različite arhitekture modela dubokog učenja za klasifikaciju poremećaja iz spektra autizma korištenjem slika funkcijske magnetske rezonancije (fMRI). Analiza je provedena na 871 uzorku iz ABIDE (Autism Brain Imaging Data Exchange) skupa podataka. Za treniranje modela korištene su matrice funkcijske povezanosti dobivene iz vremenskih serija ekstrahiranih iz 4D fMRI slika. Za model s tri skrivena sloj postignute su najveća točnost od 70,40% i preciznost od 71,55%, a najveći odziv od 77,53% i mjera F1 od 67,16% postignuti su za model s četiri skrivena sloja.

Abstract (english)

Autism spectrum disorder (ASD) is among the most prevalent neurodevelopmental disorders, and traditional diagnostic methods are often slow and subjective. In this study various deep learning model architectures were compared for classifying ASD using functional magnetic resonance imaging (fMRI) images. The analysis was conducted on 871 samples from the ABIDE (Autism Brain Imaging Data Exchange) dataset. Functional connectivity matrices derived from time series extracted from 4D fMRI images were used for training the models. The highest accuracy of 70,40% and precision of 71,55% were achieved by a model with three hidden layers, while the highest recall of 77,53% and F1-score of 67,16% were achieved by a model with four hidden layers.

• Keywords: duboko učenje; klasifikacije; autizam; fMRI; ABIDE
• Keywords (english): deep learning; classification; ASD; fMRI; ABIDE
• Language: croatian
• URN:NBN: urn:nbn:hr:168:867009
• 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 21:46:37