Page 94 - AN-4-4

Advanced Neurology

ORIGINAL RESEARCH ARTICLE
Enhancing brain tumor classification with a

diffusion denoising model and a conditional
deep convolutional neural network

Efe Precious Onakpojeruo 1,2,3 * , Dilber Uzun Ozsahin 1,4,5 , and Ilker Ozsahin 1,6

1 Operational Research Centre in Healthcare, Near East University, TRNC Mersin 10, Nicosia, Turkey
2 Department of Integrated Biomedical Graduate Studies, School of Medicine, Loma Linda University,
California, United States of America
3 Department of Biomedical Engineering, Faculty of Engineering, Near East University, TRNC Mersin
10, Nicosia, Turkey
4 Department of Medical Diagnostic Imaging, College of Health Sciences, University of Sharjah,
Sharjah, United Arab Emirates
5 Research Institute for Medical and Health Sciences, University of Sharjah, Sharjah, United Arab Emirates
6 Department of Mathematical Sciences, Saveetha School of Engineering, Saveetha Institute of
Medical and Technical Sciences, Chennai, Tamil Nadu, India

Abstract
*Corresponding author:
Efe Precious Onakpojeruo The limited availability of medical imaging datasets and concerns over patient
(efeprecious.onakpojeruo@neu. privacy pose significant challenges in artificial intelligence-driven disease
edu.tr) diagnosis. To overcome these limitations, this study introduces the use of the
Citation: Onakpojeruo EP, denoising diffusion model (DDM) for generating synthetic datasets, marking
Ozsahin DU, Ozsahin I. Enhancing a significant advancement over traditional generative adversarial networks
brain tumor classification with a
diffusion denoising model and a (GANs). This research pioneers the integration of DDM with conditional deep
conditional deep convolutional convolutional neural networks (CDCNN) for brain tumor classification, focusing
neural network. Adv Neurol. on four categories: Glioma, meningioma, pituitary tumors, and healthy tissue. The
2025;4(4):88-100.
doi: 10.36922/AN025130025 proposed CDCNN model, developed from existing convolutional neural network
architectures, effectively processed both DDM-generated synthetic datasets and
Received: March 24, 2025
original datasets sourced from the Kaggle repository. The results demonstrate the
1st revised: July 15, 2025 remarkable efficacy of the DDM-based augmentation framework, with the CDCNN
2nd revised: August 19, 2025 model achieving an accuracy of 96.2%, significantly outperforming traditional
GAN-based models, such as Pix2Pix. A comparative analysis against established
Accepted: August 28, 2025
architectures, including ResNet50, Visual Geometry Group (VGG)16, VGG19, and
Published online: September 17, InceptionV3, further highlights the superior sensitivity, specificity, and F1 score of
2025 the proposed framework. These findings underscore the transformative potential
Copyright: © 2025 Author(s). of diffusion models in enhancing dataset diversity, improving classification
This is an Open-Access article performance, and addressing data scarcity issues in medical imaging. The
distributed under the terms of the
Creative Commons Attribution proposed framework offers a scalable, robust solution for brain tumor diagnosis,
License, permitting distribution, paving the way for improved disease prediction and treatment planning in clinical
and reproduction in any medium, practice.
provided the original work is
properly cited.
Publisher’s Note: AccScience Keywords: Brain tumors; Conditional deep convolutional neural network; Denoising
Publishing remains neutral with diffusion model; Synthetic data augmentation; Medical image classification; Generative
regard to jurisdictional claims in
published maps and institutional adversarial networks
affiliations.

Volume 4 Issue 4 (2025) 88 doi: 10.36922/AN025130025