ML Security PapersDiffusion-Based Feature Denoising and Using NNMF for Robust Brain Tumor Classification
Hiba Adil Al-kharsan 1, Róbert Rajkó 1,2
Published on arXiv
2603.13182
Input Manipulation Attack
OWASP ML Top 10 — ML01
Key Finding
Achieves competitive classification accuracy on brain tumor MRI while significantly enhancing robustness against AutoAttack adversarial perturbations through diffusion-based feature denoising
NNMF-Diffusion Defense
Novel technique introduced
Brain tumor classification from magnetic resonance imaging, which is also known as MRI, plays a sensitive role in computer-assisted diagnosis systems. In recent years, deep learning models have achieved high classification accuracy. However, their sensitivity to adversarial perturbations has become an important reliability concern in medical applications. This study suggests a robust brain tumor classification framework that combines Non-Negative Matrix Factorization (NNMF or NMF), lightweight convolutional neural networks (CNNs), and diffusion-based feature purification. Initially, MRI images are preprocessed and converted into a non-negative data matrix, from which compact and interpretable NNMF feature representations are extracted. Statistical metrics, including AUC, Cohen's d, and p-values, are used to rank and choose the most discriminative components. Then, a lightweight CNN classifier is trained directly on the selected feature groups. To improve adversarial robustness, a diffusion-based feature-space purification module is introduced. A forward noise method followed by a learned denoiser network is used before classification. System performance is estimated using both clean accuracy and robust accuracy under powerful adversarial attacks created by AutoAttack. The experimental results show that the proposed framework achieves competitive classification performance while significantly enhancing robustness against adversarial perturbations.The findings presuppose that combining interpretable NNMF-based representations with a lightweight deep approach and diffusion-based defense technique supplies an effective and reliable solution for medical image classification under adversarial conditions.
Key Contributions
- Combines NNMF feature extraction with lightweight CNN for interpretable brain tumor classification
- Introduces diffusion-based feature-space purification module (forward noise + learned denoiser) to defend against adversarial perturbations
- Evaluates clean and robust accuracy against AutoAttack, demonstrating competitive performance and improved adversarial robustness
🛡️ Threat Analysis
Primary contribution is a defense against adversarial examples (AutoAttack) at inference time through diffusion-based feature purification — directly addresses input manipulation attacks on medical image classifiers.