Predikcija poremećaja iz spektra autizma korištenjem značajki strukturnih slika magnetske rezonancije mozga

undergraduate thesis

• Title (english): Autism spectrum disorder prediction using features of the structural magnetic resonance images of the brain
• Author: Lana Barić
• 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: 2024-09-11, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Autizam je kompleksan neurobiološki poremećaj ljudskog mozga koji je okarakteriziran različitim oblicima oštećenja socijalne interakcije, komunikacije i ponašanja. Kako bi osobe s autizmom dobile pravovremenu pomoć, rana dijagnoza je izrazito važna. Stoga se postavlja pitanje, može li se on dijagnosticirati i predvidjeti pomoću modela strojnog učenja te korištenjem slika magnetske rezonancije (MR) mozga. U ovom radu se fokusiralo na strukturne MR slike. Pomoću modela logističke regresije provelo se treniranje na skupu podataka ABIDE, a potom i testiranje modela. Korišteno je nekoliko skupova značajki, a najbolji rezultat je bio kada se radilo sa svim anatomskim značajkama mozga, ali bez korištenja demografskih podataka sudionika. Tada je točnost bila 60%, AUC rezultat 0,6209 te bi model logističke regresije prepoznao da osoba ima autizam u 55 % slučajeva. Dobiveni rezultati su u skladu s očekivanjima i pokazuju da bi daljnje istraživanje i razvoj modela strojnog učenja i MR slika mozga bili od velike koristi u otkrivanju autizma.

Abstract (english)

Autism is a complex neurobiological disorder of the human brain characterized by various forms of impairment in social interaction, communication, and behaviour. For individuals with autism, early diagnosis is extremely important. Therefore, the question arises whether machine learning models can be used to diagnose and predict autism with the help of brain magnetic resonance (MR) imaging. This study focused on structural MR images. A logistic regression model was trained on the ABIDE dataset and then tested. Several feature sets were used, and the best result was achieved when working with all anatomical brain features, but without using the demographic data of the participants. In that case, accuracy was 60%, the AUC score was 0.6209, and the logistic regression model would recognize that a person has autism in 55% of cases. The results obtained are in line with expectations and indicate that further research and development of machine learning models and brain MR images would be of great benefit in detection of autism.

• Keywords: autizam; predikcija; strojno učenje; strukturne slike magnetske rezonancije mozga
• Keywords (english): autism; prediction; machine learning; structural MR images
• Language: croatian
• URN:NBN: urn:nbn:hr:168:249974
• 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 20:03:36