Residual RAKI: A hybrid linear and non-linear approach for scan-specific k-space deep learning

Zhang, Chi; Moeller, Steen; Demirel, Ömer Burak; Uğurbil, Kâmil; Akçakaya, Mehmet

doi:10.1016/j.neuroimage.2022.119248

Cited by 10 publications

(10 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…41 Leveraging the nonlinear modeling capabilities of deep neural networks, some researchers have started applying DL to k-space-based PI. [61][62][63][64][65] Compared to traditional linear interpolation methods, the performance of nonlinear interpolation methods with the assistance of deep neural networks is superior, and the models are more robust. [61][62][63][64][65] More information on DL for parallel MR imaging can be referred to this review.…”

Section: Pimentioning

confidence: 99%

Knowledge‐driven deep learning for fast MR imaging: Undersampled MR image reconstruction from supervised to un‐supervised learning

Wang,

Wu,

Jia

et al. 2024

Magnetic Resonance in Med

View full text Add to dashboard Cite

Deep learning (DL) has emerged as a leading approach in accelerating MRI. It employs deep neural networks to extract knowledge from available datasets and then applies the trained networks to reconstruct accurate images from limited measurements. Unlike natural image restoration problems, MRI involves physics‐based imaging processes, unique data properties, and diverse imaging tasks. This domain knowledge needs to be integrated with data‐driven approaches. Our review will introduce the significant challenges faced by such knowledge‐driven DL approaches in the context of fast MRI along with several notable solutions, which include learning neural networks and addressing different imaging application scenarios. The traits and trends of these techniques have also been given which have shifted from supervised learning to semi‐supervised learning, and finally, to unsupervised learning methods. In addition, MR vendors' choices of DL reconstruction have been provided along with some discussions on open questions and future directions, which are critical for the reliable imaging systems.

show abstract

Section: Pimentioning

confidence: 99%

Knowledge‐driven deep learning for fast MR imaging: Undersampled MR image reconstruction from supervised to un‐supervised learning

Wang,

Wu,

Jia

et al. 2024

Magnetic Resonance in Med

View full text Add to dashboard Cite

show abstract

“…The scope of this work was to improve RAKI by iterative training. Another approach for improving RAKI includes residual RAKI 6 (rRAKI) that trains non‐linear CNNs jointly with a linear convolution implemented via a skip connection. In our experience, however, rRAKI suffers from similar limitations regarding the amount of training data.…”

Section: Discussionmentioning

confidence: 99%

“…Similar to GRAPPA, the neural network parameters in RAKI (i.e., the convolution filter weights within the CNN) are calibrated using scan‐specific ACS as training data. In previous studies, RAKI has demonstrated better performance in comparison to GRAPPA 4–6 …”

Section: Introductionmentioning

confidence: 93%

“…In previous studies, RAKI has demonstrated better performance in comparison to GRAPPA. [4][5][6] However, RAKI requires more training data because of its increased parameter space, which limits its applicability in standard 2D imaging, because the effective acceleration is significantly decreased. For small matrix sizes and high acceleration factors, the acquisition of the training data may take as long as the acquisition of the undersampled imaging data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Iterative training of robust k‐space interpolation networks for improved image reconstruction with limited scan specific training samples

Dawood

Breuer

Stebani

et al. 2022

Magnetic Resonance in Med

View full text Add to dashboard Cite

Purpose: To evaluate an iterative learning approach for enhanced performance of robust artificial-neural-networks for k-space interpolation (RAKI), when only a limited amount of training data (auto-calibration signals [ACS]) are available for accelerated standard 2D imaging. Methods:In a first step, the RAKI model was tailored for the case of limited training data amount. In the iterative learning approach (termed iterative RAKI [iRAKI]), the tailored RAKI model is initially trained using original and augmented ACS obtained from a linear parallel imaging reconstruction. Subsequently, the RAKI convolution filters are refined iteratively using original and augmented ACS extracted from the previous RAKI reconstruction. Evaluation was carried out on 200 retrospectively undersampled in vivo datasets from the fastMRI neuro database with different contrast settings.Results: For limited training data (18 and 22 ACS lines for R = 4 and R = 5, respectively), iRAKI outperforms standard RAKI by reducing residual artifacts and yields better noise suppression when compared to standard parallel imaging, underlined by quantitative reconstruction quality metrics. Additionally, iRAKI shows better performance than both GRAPPA and standard RAKI in case of pre-scan calibration with varying contrast between training-and undersampled data. Conclusion:RAKI benefits from the iterative learning approach, which preserves the noise suppression feature, but requires less original training data for the accurate reconstruction of standard 2D images thereby improving net acceleration.

show abstract

“…Recently, CS has been successfully applied to the BB imaging of carotid plaque, where it exhibited a comparable image quality to the fully sampled methods with significantly reduced scanner times (27)(28)(29). However, the image quality contains blurring, and the image contrast appears reduced for high accelerations when the conventional CS is used, such as the ℓ 1 -SPIRiT regularization (30).…”

Section: Introductionmentioning

confidence: 99%

High-efficiency 3D black-blood thoracic aorta imaging with patch-based low-rank tensor reconstruction

Shi¹,

Liu²,

Cheng³

et al. 2023

Quant Imaging Med Surg

View full text Add to dashboard Cite

Background: Three-dimensional (3D) black-blood (BB) vessel wall imaging is a promising noninvasive imaging technique for assessing thoracic aortic diseases. We aimed to develop and evaluate a fast thoracic aorta vessel wall imaging method with patch-based low-rank tensor (Pt-LRT) reconstruction using the 3D-modulated variable flip angle fast-spin echo (vFA-FSE) sequence. Methods:The Pt-LRT technique adopts a low-rank tensor image model with regularization to explore the local low-rankness and nonlocal redundancies of the images to assess the thoracic aorta vessel wall. It uses high-order tensors to capture correlations between data in multiple dimensions and reconstructs images from highly undersampled data. For this study, 12 healthy participants and 2 patients with thoracic aortic diseases were evaluated at 3T magnetic resonance (MR). The reconstruction results were compared to the traditional generalized autocalibrating partially parallel acquisitions (GRAPPA) and ℓ 1 -SPIRiT reconstruction to assess the feasibility of the proposed framework. Quantitative analyses of the vessel wall thickness (VWT), internal diameter (ID), lumen area (LA), and contrast-to-noise ratio (CNR) between the lumen and vessel wall were performed on all healthy participants.Results: Results demonstrated no significant differences between the GRAPPA and the proposed Pt-LRT in VWT, ID, or LA of the aorta (P<0.05). A higher mean CNR was attained with 3D patch-based low-rank tensor reconstruction than with ℓ 1 -SPIRiT reconstruction (49.4±10.8 vs. 38.9±8.2). Conclusions:The proposed 3D BB thoracic aorta vessel wall imaging method can reduce the scan time and produce an image quality that is in good agreement with the conventional GRAPPA acquisition, which takes approximately more than 8 min. This study also shows that the proposed Pt-LRT method substantially improves the visualization and sharpness of the vessel wall and the definition of the tissue boundary compared to the imaging obtained with ℓ 1 -SPIRiT.

show abstract

Residual RAKI: A hybrid linear and non-linear approach for scan-specific k-space deep learning

Cited by 10 publications

References 53 publications

Knowledge‐driven deep learning for fast MR imaging: Undersampled MR image reconstruction from supervised to un‐supervised learning

Knowledge‐driven deep learning for fast MR imaging: Undersampled MR image reconstruction from supervised to un‐supervised learning

Iterative training of robust k‐space interpolation networks for improved image reconstruction with limited scan specific training samples

High-efficiency 3D black-blood thoracic aorta imaging with patch-based low-rank tensor reconstruction

Contact Info

Product

Resources

About