An Unsupervised Approach for Artifact Severity Scoring in Multi-Contrast MR Images

Savannah Hays, Lianrui Zuo, Blake E. Dewey, Samuel Remedios, Jinwei Zhang, Ellen M. Mowry, Scott D. Newsome, Aaron Carass, Jerry L Prince
Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, PMLR 301:604-614, 2026.

Abstract

Quality assurance (QA) in magnetic resonance (MR) imaging is critical but remains a challenging and time-intensive process, particularly when working with large-scale, multi-site imaging datasets. Manual QA methods are subjective, prone to inter-rater variability, and impractical for high-throughput workflows. Existing automated QA methods often lack generalizability to diverse datasets or fail to provide interpretable insights into the causes of poor image quality. To address these limitations, we introduce an unsupervised and interpretable QA framework for multi-contrast MR images that quantifies artifact severity. By assigning a numerical score to each image, our method enables objective, consistent evaluation of image quality and highlights specific levels of artifact presence that can impair downstream analysis. Our framework employs an unsupervised contrastive learning approach, leveraging simulated artifact transformations, including random bias, noise, anisotropy, and ghosting, to train the model without requiring manual labels or preprocessing. A margin-based contrastive loss further enables differentiation between varying levels of artifact severity. We validate our framework using simulated artifacts on a public dataset and real artifacts on a private clinical dataset, demonstrating its robustness and generalizability for automatic MR image QA. By efficiently evaluating image quality and identifying artifacts prior to data processing, our approach streamlines QA workflows and enhances the reliability of subsequent analyses in both research and clinical settings.

Cite this Paper

BibTeX
@InProceedings{pmlr-v301-hays26a, title = {An Unsupervised Approach for Artifact Severity Scoring in Multi-Contrast MR Images}, author = {Hays, Savannah and Zuo, Lianrui and Dewey, Blake E. and Remedios, Samuel and Zhang, Jinwei and Mowry, Ellen M. and Newsome, Scott D. and Carass, Aaron and Prince, Jerry L}, booktitle = {Proceedings of The 8th International Conference on Medical Imaging with Deep Learning}, pages = {604--614}, year = {2026}, editor = {Tasdizen, Tolga and Elhabian, Shireen and Summers, Ronald and Chen, Chen and Koch, Lisa and Zhuang, Yan}, volume = {301}, series = {Proceedings of Machine Learning Research}, month = {09--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v301/main/assets/hays26a/hays26a.pdf}, url = {https://proceedings.mlr.press/v301/hays26a.html}, abstract = {Quality assurance (QA) in magnetic resonance (MR) imaging is critical but remains a challenging and time-intensive process, particularly when working with large-scale, multi-site imaging datasets. Manual QA methods are subjective, prone to inter-rater variability, and impractical for high-throughput workflows. Existing automated QA methods often lack generalizability to diverse datasets or fail to provide interpretable insights into the causes of poor image quality. To address these limitations, we introduce an unsupervised and interpretable QA framework for multi-contrast MR images that quantifies artifact severity. By assigning a numerical score to each image, our method enables objective, consistent evaluation of image quality and highlights specific levels of artifact presence that can impair downstream analysis. Our framework employs an unsupervised contrastive learning approach, leveraging simulated artifact transformations, including random bias, noise, anisotropy, and ghosting, to train the model without requiring manual labels or preprocessing. A margin-based contrastive loss further enables differentiation between varying levels of artifact severity. We validate our framework using simulated artifacts on a public dataset and real artifacts on a private clinical dataset, demonstrating its robustness and generalizability for automatic MR image QA. By efficiently evaluating image quality and identifying artifacts prior to data processing, our approach streamlines QA workflows and enhances the reliability of subsequent analyses in both research and clinical settings.} }
Endnote
%0 Conference Paper %T An Unsupervised Approach for Artifact Severity Scoring in Multi-Contrast MR Images %A Savannah Hays %A Lianrui Zuo %A Blake E. Dewey %A Samuel Remedios %A Jinwei Zhang %A Ellen M. Mowry %A Scott D. Newsome %A Aaron Carass %A Jerry L Prince %B Proceedings of The 8th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Tolga Tasdizen %E Shireen Elhabian %E Ronald Summers %E Chen Chen %E Lisa Koch %E Yan Zhuang %F pmlr-v301-hays26a %I PMLR %P 604--614 %U https://proceedings.mlr.press/v301/hays26a.html %V 301 %X Quality assurance (QA) in magnetic resonance (MR) imaging is critical but remains a challenging and time-intensive process, particularly when working with large-scale, multi-site imaging datasets. Manual QA methods are subjective, prone to inter-rater variability, and impractical for high-throughput workflows. Existing automated QA methods often lack generalizability to diverse datasets or fail to provide interpretable insights into the causes of poor image quality. To address these limitations, we introduce an unsupervised and interpretable QA framework for multi-contrast MR images that quantifies artifact severity. By assigning a numerical score to each image, our method enables objective, consistent evaluation of image quality and highlights specific levels of artifact presence that can impair downstream analysis. Our framework employs an unsupervised contrastive learning approach, leveraging simulated artifact transformations, including random bias, noise, anisotropy, and ghosting, to train the model without requiring manual labels or preprocessing. A margin-based contrastive loss further enables differentiation between varying levels of artifact severity. We validate our framework using simulated artifacts on a public dataset and real artifacts on a private clinical dataset, demonstrating its robustness and generalizability for automatic MR image QA. By efficiently evaluating image quality and identifying artifacts prior to data processing, our approach streamlines QA workflows and enhances the reliability of subsequent analyses in both research and clinical settings.
APA
Hays, S., Zuo, L., Dewey, B.E., Remedios, S., Zhang, J., Mowry, E.M., Newsome, S.D., Carass, A. & Prince, J.L.. (2026). An Unsupervised Approach for Artifact Severity Scoring in Multi-Contrast MR Images. Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 301:604-614 Available from https://proceedings.mlr.press/v301/hays26a.html.

Related Material