Home Publications Research Grants Inventions & Patents Awards Additional Research Activities Faculties & Institutions Research Areas

Deep learning based denoising of mammographic x-ray images: an investigation of loss functions and their detail-preserving properties

Eckert D, Ritschl L, Herbst M, Wicklein J, Vesal S, Kappler S, Maier A, Stober S (2022)

---

Publication Type: Conference contribution

Publication year: 2022

Journal

Publisher: SPIE

Book Volume: 12031

Conference Proceedings Title: Progress in Biomedical Optics and Imaging - Proceedings of SPIE

Event location: Virtual, Online

ISBN: 9781510649378

DOI: 10.1117/12.2612403

Abstract

Digital Breast Tomosynthesis (DBT) is becoming increasingly popular for breast cancer screening because of its high depth resolution. It uses a set of low-dose x-ray images called raw projections to reconstruct an arbitrary number of planes. These are typically used in further processing steps like backprojection to generate DBT slices or synthetic mammography images. Because of their low x-ray dose, a high amount of noise is present in the projections. In this study, the possibility of using deep learning for the removal of noise in raw projections is investigated. The impact of loss functions on the detail preservation is analized in particular. For that purpose, training data is augmented following the physics driven approach of Eckert et al.1 In this method, an x-ray dose reduction is simulated. First pixel intensities are converted to the number of photons at the detector. Secondly, Poisson noise is enhanced in the x-ray image by simulating a decrease in the mean photon arrival rate. The Anscombe Transformation2 is then applied to construct signal independent white Gaussian noise. The augmented data is then used to train a neural network to estimate the noise. For training several loss functions are considered including the mean square error (MSE), the structural similarity index (SSIM)3 and the perceptual loss.4 Furthermore the ReLU-Loss1 is investigated, which is especially designed for mammogram denoising and prevents the network from noise overestimation. The denoising performance is then compared with respect to the preservation of small microcalcifications. Based on our current measurements, we demonstrate that the ReLU-Loss in combination with SSIM improves the denoising results.

Authors with CRIS profile

Involved external institutions

How to cite

APA:

Eckert, D., Ritschl, L., Herbst, M., Wicklein, J., Vesal, S., Kappler, S.,... Stober, S. (2022). Deep learning based denoising of mammographic x-ray images: an investigation of loss functions and their detail-preserving properties. In Wei Zhao, Lifeng Yu (Eds.), Progress in Biomedical Optics and Imaging - Proceedings of SPIE. Virtual, Online: SPIE.

MLA:

Eckert, Dominik, et al. "Deep learning based denoising of mammographic x-ray images: an investigation of loss functions and their detail-preserving properties." Proceedings of the Medical Imaging 2022: Physics of Medical Imaging, Virtual, Online Ed. Wei Zhao, Lifeng Yu, SPIE, 2022.

BibTeX: Download