Deep Learning-based Denoising of Mammographic Images using Physics-driven Data Augmentation

TL;DR

This paper introduces a deep learning approach using CNNs and physics-driven data augmentation to effectively denoise mammographic images, enhancing image quality for more accurate breast cancer diagnosis.

Contribution

It presents a novel CNN-based denoising method with physics-inspired data augmentation, capable of removing both simulated and real noise in mammograms, outperforming existing techniques.

Findings

Qualitatively better denoising results than state-of-the-art methods

Effective noise removal in both simulated and real mammograms

Improved image quality for diagnostic accuracy

Abstract

Mammography is using low-energy X-rays to screen the human breast and is utilized by radiologists to detect breast cancer. Typically radiologists require a mammogram with impeccable image quality for an accurate diagnosis. In this study, we propose a deep learning method based on Convolutional Neural Networks (CNNs) for mammogram denoising to improve the image quality. We first enhance the noise level and employ Anscombe Transformation (AT) to transform Poisson noise to white Gaussian noise. With this data augmentation, a deep residual network is trained to learn the noise map of the noisy images. We show, that the proposed method can remove not only simulated but also real noise. Furthermore, we also compare our results with state-of-the-art denoising methods, such as BM3D and DNCNN. In an early investigation, we achieved qualitatively better mammogram denoising results.