Model dubokog učenja za predviđanje Alzheimerove bolesti

undergraduate thesis

• Title (english): A deep learning model to predict Alzheimer's disease
• Author: Lucian Šegota
• Mentor: Tomislav Stipančić (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2024-02-28, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

Alzheimerova bolest (AD) najrašireniji je tip demencije živčanog sustava koji uzrokuje
slabljenje mnogih moždanih funkcija (npr. gubitak pamćenja). Budući da je rana dijagnoza
najvažniji čimbenik u poboljšanju skrbi za pacijente i ishoda liječenja, neinvazivna rana
dijagnoza AD-a privukla je veliku pažnju istraživanja posljednjih godina. Cilj ovog rada je
razvoj modela temeljenog na dubokom učenju za preciznu klasifikaciju stadija AD. Koristeći
isti prošireni skup podataka, predlažu se dvije arhitekture: arhitektura konvolucijske neuronske
mreže (CNN) i arhitektura modela InceptionV3 (transferno učenje). Procjenjuju se te
arhitekture dubokog učenja, donose se zaključci iz nalaza i predlažu se poboljšanja za modele.
Skup podataka je preuzet s Kagglea koji je javno dostupan i sastoji se od četiri klase koje
predstavljaju različite faze Alzheimerove bolesti.

Abstract (english)

Alzheimer’s disease (AD) is the most prevalent type of dementia of the nervous system that
causes many brain functions to weaken (eg, memory loss). Since early diagnosis is the most
important factor in improving patient care and treatment outcomes, non-invasive early
diagnosis of AD has garnered a lot of research attention in recent years. Developing a deep
learning-based model for precise AD stage classification is the aim of this thesis. Using the
same augmented dataset, two architectures are proposed: a Convolutional Neural Network
(CNN) architecture and an InceptionV3 model architecture(transfer learning). This is to
evaluate these deep learning architectures, make inferences from the findings, and suggest
improvements for the models.. The dataset has been taken from Kaggle which is publicly
available and comprises of four classes which represents the various stages of Alzheimer's
disease.

• Keywords: Alzheimerova bolest (AD); duboko učenje; transferno učenje; MRI mozga; CNN
• Keywords (english): Alzheimer’s Disease (AD); Deep Learning; Transfer learning; Brain MRI; CNN
• Language: croatian
• URN:NBN: urn:nbn:hr:235:245413
• Study programme: Title: Mechanical Engineering; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka strojarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka strojarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2024-02-23 16:17:39