Enhancing quality and speed in database‐free neural network reconstructions of undersampled MRI with SCAMPI
Magnetic Resonance in Medicine · 2024
DOI: 10.1002/mrm.30114 ↗At a glance
Journal article in Magnetic Resonance in Medicine, 2024. Written by Thomas M Siedler and 2 co-authors. Institution: Department of Experimental Physics 5, University of Würzburg, Würzburg, Germany. Major MEDLINE topics: Magnetic Resonance Imaging, Neural Networks, Computer, Image Processing, Computer-Assisted, Algorithms. Cited 3 times so far (Crossref); 58 references. The full text is freely available (open access, CC BY 4.0). Funded by: Deutsche Forschungsgemeinschaft.
Abstract (original language)
Purpose
We present SCAMPI (Sparsity Constrained Application of deep Magnetic resonance Priors for Image reconstruction), an untrained deep Neural Network for MRI reconstruction without previous training on datasets. It expands the Deep Image Prior approach with a multidomain, sparsity‐enforcing loss function to achieve higher image quality at a faster convergence speed than previously reported methods.
Methods
Two‐dimensional MRI data from the FastMRI dataset with Cartesian undersampling in phase‐encoding direction were reconstructed for different acceleration rates for single coil and multicoil data.
Results
The performance of our architecture was compared to state‐of‐the‐art Compressed Sensing methods and ConvDecoder, another untrained Neural Network for two‐dimensional MRI reconstruction. SCAMPI outperforms these by better reducing undersampling artifacts and yielding lower error metrics in multicoil imaging. In comparison to ConvDecoder, the U‐Net architecture combined with an elaborated loss‐function allows for much faster convergence at higher image quality. SCAMPI can reconstruct multicoil data without explicit knowledge of coil sensitivity profiles. Moreover, it is a novel tool for reconstructing undersampled single coil k‐space data.
Conclusion
Our approach avoids overfitting to dataset features, that can occur in Neural Networks trained on databases, because the network parameters are tuned only on the reconstruction data. It allows better results and faster reconstruction than the baseline untrained Neural Network approach.
Subjects
MeSH terms (MEDLINE)
Author keywords
Authors and institutions
- Department of Experimental Physics 5, University of Würzburg, Würzburg, Germany.
ORCID profiles: Thomas M Siedler, Peter M Jakob, Volker Herold
Funding
- Deutsche Forschungsgemeinschaft
Cite
APA
Siedler, T. M., Jakob, P. M., & Herold, V. (2024). Enhancing quality and speed in database‐free neural network reconstructions of undersampled MRI with SCAMPI. Magnetic Resonance in Medicine. https://doi.org/10.1002/mrm.30114
Vancouver
Siedler TM, Jakob PM, Herold V. Enhancing quality and speed in database‐free neural network reconstructions of undersampled MRI with SCAMPI. Magnetic Resonance in Medicine. 2024. doi:10.1002/mrm.30114
BibTeX
@article{Siedler2024,
author = {Siedler, Thomas M and Jakob, Peter M and Herold, Volker},
title = {Enhancing quality and speed in database‐free neural network reconstructions of undersampled MRI with SCAMPI},
journal = {Magnetic Resonance in Medicine},
year = {2024},
doi = {10.1002/mrm.30114}
}
EUVIMED