Scalable Learning-Based Sampling Optimization for Compressive Dynamic MRI

Sanchez, Thomas; Gözcü, Baran; Heeswijk, Ruud B. van; Eftekhari, Armin; Ilicak, Efe; Çukur, Tolga; Cevher, Volkan

doi:10.1109/icassp40776.2020.9053345

Cited by 24 publications

(23 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Algorithm 2 is infeasible in terms running time in the case 3D MRI where one is allowed to sample in both phase and frequency encoding directions. A faster stochastic approach was proposed in [24] for dynamic MRI, but cannot be readily applied to the 3D setting, as it would require unreasonably large for S ∈ S such that c(Ω ∪ S) ≤ Γ do 4:…”

Section: Learning-based Lazy Greedy Algorithmmentioning

confidence: 99%

Rethinking Sampling in Parallel MRI: A Data-Driven Approach

Gözcü¹,

Sanchez²,

Cevher³

2019

2019 27th European Signal Processing Conference (EUSIPCO)

Self Cite

View full text Add to dashboard Cite

In the last decade, Compressive Sensing (CS) has emerged as the most promising, model-driven approach to accelerate MRI scans. CS relies on the key sparsity assumption and proposes random sampling for data acquisition. The practical CS approaches in MRI employ variable-density (VD) sampling, where samples are drawn at random based on a parametric probability model which focuses on the center of the Fourier domain. In stark contrast to this model-driven sampling approaches, we propose a data-driven framework for optimizing sampling in parallel (multi-coil) MRI. Our approach does not assume any structure in the data, and instead optimizes a performance metric (e.g. PSNR) for any given reconstruction algorithm, based on our earlier learning-based sampling framework previously applied to 2D MRI which we also extend to 3D MRI setting in this work by employing lazy evaluations in the greedy algorithm. We show boosted performance for the parallel MRI based on this sampling approach and highlight the inefficiency of variable density approaches. This suggests that data-driven sampling methods could be the key to unlocking the full power of CS applied to MRI.

show abstract

Section: Learning-based Lazy Greedy Algorithmmentioning

confidence: 99%

Rethinking Sampling in Parallel MRI: A Data-Driven Approach

Gözcü¹,

Sanchez²,

Cevher³

2019

2019 27th European Signal Processing Conference (EUSIPCO)

Self Cite

View full text Add to dashboard Cite

show abstract

“…The comparison is based on the performance of FISTA when reconstructing the underlying reflectivity x in a ROI at a fixed compression rate. This means that N F , M R , and M T are specified, and compression matrices using all four combinations of spatial and frequency subsampling approaches are constructed following (20). Additionally, sample reconstructions of measurement data are provided for qualitative evaluation.…”

Section: Methodsmentioning

confidence: 99%

“…Flaws beyond the horizontal aperture of the array are inherently more difficult to detect; indeed, the two leftmost and the rightmost holes are not visible when using standard TFM on the complete data set. Because of this, the maximization over x in (20)…”

Section: Simulation Scenariosmentioning

confidence: 99%

“…In this vein, co-prime arrays have been designed from closed-form expressions for Direction of Arrival (DOA) estimation paired with sparse recovery using a steering vector dictionary [ 19 ]. In [ 20 ], the compression matrix for MRI imaging is designed by estimating the best probability density function from which coefficients should be drawn. In this work, sparse array design is dealt with by studying the impact the resulting compression matrix has on the CRB of a point-like scatterer whose coordinates are to be estimated.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Subsampling Approaches for Compressed Sensing with Ultrasound Arrays in Non-Destructive Testing

Perez

Kirchhof

Krieg

et al. 2020

Sensors

View full text Add to dashboard Cite

Full Matrix Capture is a multi-channel data acquisition method which enables flexible, high resolution imaging using ultrasound arrays. However, the measurement time and data volume are increased considerably. Both of these costs can be circumvented via compressed sensing, which exploits prior knowledge of the underlying model and its sparsity to reduce the amount of data needed to produce a high resolution image. In order to design compression matrices that are physically realizable without sophisticated hardware constraints, structured subsampling patterns are designed and evaluated in this work. The design is based on the analysis of the Cramér–Rao Bound of a single scatterer in a homogeneous, isotropic medium. A numerical comparison of the point spread functions obtained with different compression matrices and the Fast Iterative Shrinkage/Thresholding Algorithm shows that the best performance is achieved when each transmit event can use a different subset of receiving elements and each receiving element uses a different section of the echo signal spectrum. Such a design has the advantage of outperforming other structured patterns to the extent that suboptimal selection matrices provide a good performance and can be efficiently computed with greedy approaches.

show abstract

“…Others have tried to learn the under-sampling strategy in a supervised way. In [40], the under-sampling strategy is learned with a greedy optimization. In [41], a gradient descent optimization is used.…”

Section: Under-samplingmentioning

confidence: 99%

Benchmarking MRI Reconstruction Neural Networks on Large Public Datasets

2020

View full text Add to dashboard Cite

Deep learning is starting to offer promising results for reconstruction in Magnetic Resonance Imaging (MRI). A lot of networks are being developed, but the comparisons remain hard because the frameworks used are not the same among studies, the networks are not properly re-trained, and the datasets used are not the same among comparisons. The recent release of a public dataset, fastMRI, consisting of raw k-space data, encouraged us to write a consistent benchmark of several deep neural networks for MR image reconstruction. This paper shows the results obtained for this benchmark, allowing to compare the networks, and links the open source implementation of all these networks in Keras. The main finding of this benchmark is that it is beneficial to perform more iterations between the image and the measurement spaces compared to having a deeper per-space network.

show abstract

Scalable Learning-Based Sampling Optimization for Compressive Dynamic MRI

Cited by 24 publications

References 43 publications

Rethinking Sampling in Parallel MRI: A Data-Driven Approach

Rethinking Sampling in Parallel MRI: A Data-Driven Approach

Subsampling Approaches for Compressed Sensing with Ultrasound Arrays in Non-Destructive Testing

Benchmarking MRI Reconstruction Neural Networks on Large Public Datasets

Contact Info

Product

Resources

About