Enhanced Classification of Alzheimer’s Disease Stages Using EfficientNet and ResNet50 on MRI Scans
Dasinya Journal for Engineering and Informatics · 16.07.2026 · 2 (3)
DOI: 10.65542/djei.v2i3.56 ↗Kurz gefasst
Fachartikel in der Zeitschrift Dasinya Journal for Engineering and Informatics, online erschienen am 16.07.2026. Verfasst von Rozin Majeed Abdullah und 1 weiteren Autor:innen. Die Arbeit nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Themen: Demenz
Abstract (Originalsprache)
Correct characterization of dementia stages based on magnetic resonance imaging (MRI) scans is essential for early diagnosis and treatment planning. In this work, a soft-voting ensemble model is proposed that combines EfficientNet-B3 and ResNet50 to improve classification performance across four dementia classes, namely Non-Demented, Very Mild Demented, Mild Demented, and Moderate Demented. The two baseline models are trained separately, and their outputs are subsequently averaged to generate the final prediction of the ensemble model. The experimental results demonstrate that the ensemble model achieves significantly higher performce than the individual models. The recall and F1-score reach 97.2% and 97.0%, respectively, representing an improvement of approximately 2–3% over the baseline models. The results also show smooth convergence during training, with minimal overfitting, as evidenced by the training and validation curves. In addition, the confusion matrix analysis confirms superior performance across all dementia classes. Overall, the findings indicate that ensemble learning is an effective approach for improving medical image classification tasks. Furthermore, it provides a reliable framework for dementia stage detection using MRI scans.
Zitieren
APA
Abdullah, R. M., & Abdulkarim, N. S. (2026). Enhanced Classification of Alzheimer’s Disease Stages Using EfficientNet and ResNet50 on MRI Scans. Dasinya Journal for Engineering and Informatics, 2(3). https://doi.org/10.65542/djei.v2i3.56
Vancouver
Abdullah RM, Abdulkarim NS. Enhanced Classification of Alzheimer’s Disease Stages Using EfficientNet and ResNet50 on MRI Scans. Dasinya Journal for Engineering and Informatics. 2026;2(3). doi:10.65542/djei.v2i3.56
BibTeX
@article{Abdullah2026,
author = {Abdullah, Rozin Majeed and Abdulkarim, Nashat Salih},
title = {Enhanced Classification of Alzheimer’s Disease Stages Using EfficientNet and ResNet50 on MRI Scans},
journal = {Dasinya Journal for Engineering and Informatics},
year = {2026},
volume = {2},
number = {3},
doi = {10.65542/djei.v2i3.56}
}
EUVIMED