The DOI matches its Crossref record; 18 of the 19 references it cites were found in a registry or an index.
A Bias-Reducing Loss Function for CT Image Denoising
Abstract
There is growing interest in the use of deep neural network (DNN) based image denoising to reduce patient’s X-ray dosage in medical computed tomography (CT). An effective denoiser must remove noise while maintaining the texture and detail. Commonly used mean squared error (MSE) loss functions in the DNN training weight errors due to bias and variance equally. However, the error due to bias is often more egregious since it results in loss of image texture and detail. In this paper, we present a novel approach to designing a loss function that penalizes variance and bias differently. Our proposed bias-reducing loss function allows us to train a DNN denoiser so that the amount of texture and detail retained can be controlled through a user adjustable parameter. Our experiments verify that the proposed loss function enhances the texture and detail in denoised images with only a slight increase in the MSE.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“In addition, loss function such as mean square error introduces bias in the training weights impacting the image feature quality and affecting the prediction [94].”
From Artificial intelligence bias in medical system designs: a systematic review (2023) · cites this paper as [94]
Citations
Cites
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 1×
“To train the network, we used the Adam optimizer [18] with an initial learning rate of 0.”
MethodImage quality assessment: from error visibility to structural similarity
Wang, Z., Bovik, A. C., Sheikh, H. R., & Simoncelli, E. P. (2004) · IEEE Transactions on Image Processing · doi:10.1109/tip.2003.819861 · Cited by 57,892 · cited 1×
“The similarity of denoised images with ground truth images was quantified using the average SSIM (structural similarity) [20] and PSNR.”
MethodImage denoising by sparse 3-D transform-domain collaborative filtering
Dabov, K., Foi, A., Katkovnik, V., & Egiazarian, K. (2007) · IEEE Transactions on Image Processing · doi:10.1109/tip.2007.901238 · Cited by 9,283 · cited 1×
“An alternative approach is to apply denoising in either the sinogram [4, 5] or space domain [6, 7].”
MethodBeyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
Zhang, K., Zuo, W., Chen, Y., Meng, D., & Zhang, L. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2662206 · Cited by 8,983 · cited 1×
“To do this, we used the architecture of a CNN denoiser adopted from [16] with a single input channel and 17 convolution layers.”
MethodA Non-Local Algorithm for Image Denoising
Buades, A., Coll, B., & Morel, J.-M. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.38 · Cited by 7,616 · cited 1×
“An alternative approach is to apply denoising in either the sinogram [4, 5] or space domain [6, 7].”
MethodSignature Verification Using A "Siamese" Time Delay Neural Network
BROMLEY, J., BENTZ, J. W., BOTTOU, L., GUYON, I., LECUN, Y., MOORE, C., SÄCKINGER, E., & SHAH, R. (1993) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s0218001493000339 · Cited by 2,077 · cited 1×
“Furthermore, each of the two denoising networks share the same parameters, so this is treated as a Siamese network [15] for training.”
BackgroundLow-Dose CT With a Residual Encoder-Decoder Convolutional Neural Network
Chen, H., Zhang, Y., Kalra, M. K., Lin, F., Chen, Y., Liao, P., Zhou, J., & Wang, G. (2017) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2017.2715284 · Cited by 1,866 · cited 1×
“In particular, the mean squared error (MSE) loss function is commonly used for training [10] because it results in a trained network that approximately maximizes the peak signal-to-noise ratio (PSNR).”
MethodYang, Q., Yan, P., Zhang, Y., Yu, H., Shi, Y., Mou, X., Kalra, M. K., Zhang, Y., Sun, L., & Wang, G. (2018) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2018.2827462 · Cited by 1,630 · cited 1×
“The GAN attempts to produce a denoised image with the same distribution as ground truth [12, 13].”
BackgroundComputed Tomography: Principles, Design, Artifacts, and Recent Advances
Hsieh, J. (2015) · doi:10.1117/3.2197756 · Cited by 1,400 · cited 1×
“X-ray computed tomography (CT) is perhaps the most widely used 3D medical imaging modality, and over the past decade, there has been a great deal of progress in the development of methods to further reduce noise and artifacts while improving resolution and quality [1].”
BackgroundGenerative Adversarial Networks for Noise Reduction in Low-Dose CT
Wolterink, J. M., Leiner, T., Viergever, M. A., & Isgum, I. (2017) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2017.2708987 · Cited by 1,037 · cited 1×
“However, it is also known that the MSE loss function tends to produce images that are overly smooth and lack texture [11].”
Background
Research works printed but not linked: 2
- bouman, 2013, Model Based in Image Processing
- hsieh, 2019, A new era of image reconstruction: TruefidelityTM, Technical white paper on deep learning image reconstruction
Cited by
Gao, Q., Li, Z., Zhang, J., Zhang, Y., & Shan, H. (2024) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3320812 · Cited by 162
“Despite the outstanding denoising performance, these methods often lead to over-smoothing images [16], [17].”
BackgroundArtificial intelligence bias in medical system designs: a systematic review
Kumar, A., Aelgani, V., Vohra, R., Gupta, S. K., Bhagawati, M., Paul, S., Saba, L., Suri, N., Khanna, N. N., Laird, J. R., et al. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-16029-x · Cited by 65
CoCoDiff: a contextual conditional diffusion model for low-dose CT image denoising
Gao, Q., & Shan, H. (2022) · Developments in X-Ray Tomography · doi:10.1117/12.2634939 · Cited by 59
Speckle2Self: Self-supervised ultrasound speckle reduction without clean data
Li, X., Navab, N., & Jiang, Z. (2025) · Medical Image Analysis · doi:10.1016/j.media.2025.103755 · Cited by 33
“As noted in prior work Demir et al. (2025); Nagare et al. (2021), the use of self-supervised learning frameworks and MSE loss functions may also result in relatively smooth outputs.”
Shan, H., Vimieiro, R. D. B., Borges, L. R. A. P., Vieira, M. A. C., & Wang, G. (2023) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2023.102555 · Cited by 20
“This was demonstrated in [47], where the authors proposed a loss function that takes into account the bias and noise variance for the restoration algorithm.”
Background
Metadata
- Authors
- Nagare, M., Melnyk, R., Rahman, O., Sauer, K. D., & Bouman, C. A.
- Affiliations
- Purdue University West Lafayette (US)
- University of Notre Dame (US)
- Venue
- ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
- Publisher
- IEEE
- Published
- 2021-06-06
- Pages
- 1175-1179
- ISSN
- 1520-6149 (ISSN-L)
- OpenAlex
- W3162819150
- Cited by
- 20
- References
- 19
- Citations by year
- 2026: 2 · 2025: 8 · 2023: 9 · 2022: 1
- FWCI
- 0.82
- Type
- conference paper
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE International Conference on Acoustics Speech and Signal Processing (publishedVersion)
- Funders
- GE Healthcare
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Signal Denoising Methods
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Enhancement Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Video Quality Assessment
- Keywords
- CT image denoising, bias-variance trade-off, texture preservation, X-ray dose reduction, mean squared error loss
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: A Bias-Reducing Loss Function for CT Image Denoising. https://app.cochat.ai/papers/10.1109/icassp39728.2021.9413855
- The paper
- Nagare, M., Melnyk, R., Rahman, O., Sauer, K. D., & Bouman, C. A. (2021). A Bias-Reducing Loss Function for CT Image Denoising. ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). https://doi.org/10.1109/icassp39728.2021.9413855
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.