Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 May 17.
doi: 10.1007/s00330-025-11670-6. Online ahead of print.

AI in motion: the impact of data augmentation strategies on mitigating MRI motion artifacts

Affiliations

AI in motion: the impact of data augmentation strategies on mitigating MRI motion artifacts

Simon D Westfechtel et al. Eur Radiol. .

Abstract

Objectives: Artifacts in clinical MRI can compromise the performance of AI models. This study evaluates how different data augmentation strategies affect an AI model's segmentation performance under variable artifact severity.

Materials and methods: We used an AI model based on the nnU-Net architecture to automatically quantify lower limb alignment using axial T2-weighted MR images. Three versions of the AI model were trained with different augmentation strategies: (1) no augmentation ("baseline"), (2) standard nnU-net augmentations ("default"), and (3) "default" plus augmentations that emulate MR artifacts ("MRI-specific"). Model performance was tested on 600 MR image stacks (right and left; hip, knee, and ankle) from 20 healthy participants (mean age, 23 ± 3 years, 17 men), each imaged five times under standardized motion to induce artifacts. Two radiologists graded each stack's artifact severity as none, mild, moderate, and severe, and manually measured torsional angles. Segmentation quality was assessed using the Dice similarity coefficient (DSC), while torsional angles were compared between manual and automatic measurements using mean absolute deviation (MAD), intraclass correlation coefficient (ICC), and Pearson's correlation coefficient (r). Statistical analysis included parametric tests and a Linear Mixed-Effects Model.

Results: MRI-specific augmentation resulted in slightly (yet not significantly) better performance than the default strategy. Segmentation quality decreased with increasing artifact severity, which was partially mitigated by default and MRI-specific augmentations (e.g., severe artifacts, proximal femur: DSCbaseline = 0.58 ± 0.22; DSCdefault = 0.72 ± 0.22; DSCMRI-specific = 0.79 ± 0.14 [p < 0.001]). These augmentations also maintained precise torsional angle measurements (e.g., severe artifacts, femoral torsion: MADbaseline = 20.6 ± 23.5°; MADdefault = 7.0 ± 13.0°; MADMRI-specific = 5.7 ± 9.5° [p < 0.001]; ICCbaseline = -0.10 [p = 0.63; 95% CI: -0.61 to 0.47]; ICCdefault = 0.38 [p = 0.08; -0.17 to 0.76]; ICCMRI-specific = 0.86 [p < 0.001; 0.62 to 0.95]; rbaseline = 0.58 [p < 0.001; 0.44 to 0.69]; rdefault = 0.68 [p < 0.001; 0.56 to 0.77]; rMRI-specific = 0.86 [p < 0.001; 0.81 to 0.9]).

Conclusion: Motion artifacts negatively impact AI models, but general-purpose augmentations enhance robustness effectively. MRI-specific augmentations offer minimal additional benefit.

Key points: Question Motion artifacts negatively impact the performance of diagnostic AI models for MRI, but mitigation methods remain largely unexplored. Findings Domain-specific augmentation during training can improve the robustness and performance of a model for quantifying lower limb alignment in the presence of severe artifacts. Clinical relevance Excellent robustness and accuracy are crucial for deploying diagnostic AI models in clinical practice. Including domain knowledge in model training can benefit clinical adoption.

Keywords: Artifacts; Artificial intelligence; Lower limbs; Magnetic resonance imaging; Torsion abnormality.

PubMed Disclaimer

Conflict of interest statement

Compliance with ethical standards. Guarantor: The scientific guarantor of this publication is Simon Westfechtel. Conflict of interest: D.T. received honoraria for lectures by Bayer, GE, and Philips and holds shares in StratifAI GmbH, Germany, and in Synagen GmbH, Germany, neither of whom have supported or influenced this study. All ethical standards have been strictly adhered to. Statistics and biometry: One of the authors has significant statistical expertise. Informed consent: Written informed consent was obtained from all subjects in this study. Written informed consent was waived for all patients by the Institutional Review Board. Ethical approval: Institutional Review Board approval was obtained. Study subjects or cohorts overlap: Some study subjects or cohorts have been previously reported in Schock et al [13]. Methodology: Prospective Cross-sectional study Performed at one institution

Similar articles

Cited by

References

    1. Shi Z, He L (2010) Application of neural networks in medical image processing. In: Proceedings of the second international symposium on networking and network security. Citeseer, pp 2–4
    1. Chung CB, Pathria MN, Resnick D (2024) MRI in MSK: is it the ultimate examination? Skelet Radiol 53:1727–1735. https://doi.org/10.1007/s00256-024-04601-x - DOI
    1. Budrys T, Veikutis V, Lukosevicius S et al (2018) Artifacts in magnetic resonance imaging: how it can really affect diagnostic image quality and confuse clinical diagnosis? J Vibroeng 20:1202–1213. https://doi.org/10.21595/jve.2018.19756 - DOI
    1. Singh D, Chin M, Peh W (2014) Artifacts in musculoskeletal MR imaging. Semin Musculoskelet Radiol 18:012–022. https://doi.org/10.1055/s-0034-1365831 - DOI
    1. Rafat Zand K, Reinhold C, Haider MA et al (2007) Artifacts and pitfalls in MR imaging of the pelvis. J Magn Reson Imaging 26:480–497 - DOI

LinkOut - more resources