NOise reduction with DIstribution Corrected (NORDIC) PCA in dMRI with complex-valued parameter-free locally low-rank processing

Moeller, Steen; Pisharady, Pramod Kumar; Ramanna, Sudhir; Lenglet, Christophe; Wu, Xiaoping; Dowdle, Logan T.; Yacoub, Essa; Uğurbil, Kâmil; Akçakaya, Mehmet

doi:10.1016/j.neuroimage.2020.117539

Cited by 75 publications

(111 citation statements)

References 35 publications

Supporting

Mentioning

102

Contrasting

Order By: Relevance

“…The second image reconstruction, which is the primary consideration of this manuscript, is derived from the same raw k-space files, however, following the denoising steps that aim to remove thermal noise. The denoising method, known as NORDIC for Noise Reduction with Distribution Corrected (NORDIC) PCA, is described in prior work (36, 37). In brief, this method uses a patch based, PCA implementation to identify and discard components of the data that are indistinguishable from zero-mean, normally distributed (i.e., thermal) noise, using the magnitude and complex portions of the MRI signal as input.…”

Section: Methodsmentioning

confidence: 99%

“…With the growing focus on more spatially and temporally precise fMRI measurements, we present a detailed evaluation of the achievable gains, and the resulting spatiotemporal precision, of a new denoising method, Noise Reduction with Distribution Corrected (NORDIC) PCA (36, 37). NORDIC suppresses Gaussian distributed noise associated with the MR detection process in repetitively acquired images, reducing thermal noise contributions throughout the image.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

NORDIC Increases the Sensitivity and Preserves the Spatiotemporal Precision of fMRI Responses

Dowdle

Vizioli

Moeller

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

As the neuroimaging field moves towards detecting smaller effects at higher spatial resolutions, and faster sampling rates, there is increased attention given to the deleterious contribution of unstructured, thermal noise. Here, we critically evaluate the performance of a recently developed reconstruction method, termed NORDIC, for suppressing thermal noise using datasets acquired with various field strengths, voxel sizes, sampling rates, and task designs. Following minimal preprocessing, statistical activation (t-values) of NORDIC processed data was compared to the results obtained with alternative denoising methods. Additionally, the consistency of unbiased estimations of task responses at the single-voxel, single run level, using a finite impulse response model were evaluated. To examine the potential impact on effective image resolution, the overall smoothness of the data processed with different methods was examined. Finally, to determine if NORDIC alters or removes important temporal information, an exhaustive K-fold cross validation approach was employed, using unbiased task responses to predict held out timeseries, quantified using R2. After NORDIC, the t-values are increased, an improvement comparable to what could be achieved by 1.5 voxels smoothing, and task events are clearly visible and have less cross-run error. These advantages are achieved without significant compromises in spatial and temporal resolution. Cross-validated R2s based on the unbiased models show that NORDIC is not distorting the temporal structure of the data and is the best predictor of non-denoised time courses. The results demonstrate that 1 run of NORDIC data is equivalent to using 2 to 3 original runs, and performs equally well across a diverse array of functional imaging protocols.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

NORDIC Increases the Sensitivity and Preserves the Spatiotemporal Precision of fMRI Responses

Dowdle

Vizioli

Moeller

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…First, all quantitative contrasts are generated by the same sequence and readout train, leading to similar image distortions which aid alignment of the different maps where needed. Second, for the same reason, the images from all contrasts share the same image noise distribution, which aids denoising with methods that take all images into account (Cordero-Grande et al, 2019;Moeller et al, 2021;Veraart et al, 2016). Third, well-aligned B1 + maps can help in correcting transmit field related biases in quantitative maps, for instance in qT1 maps (Marques et al, 2010).…”

Section: Quantitative Multi-contrast Mappingmentioning

confidence: 99%

MESMERISED: Super-accelerating T₁relaxometry and diffusion MRI with STEAM at 7 T for quantitative multi-contrast and diffusion imaging

Fritz

Poser²,

Roebroeck³

2020

Preprint

View full text Add to dashboard Cite

There is an increasing interest in quantitative imaging of T1, T2 and diffusion contrast in the brain due to greater robustness against bias fields and artifacts, as well as better biophysical interpretability in terms of microstructure. However, acquisition time constraints are a challenge, particularly when multiple quantitative contrasts are desired and when extensive sampling of diffusion directions, high b-values or long diffusion times are needed for multi-compartment microstructure modeling. Although ultra-high fields of 7 T and above have desirable properties for many MR modalities, the shortening T2 and the high specific absorption rate (SAR) of inversion and refocusing pulses bring great challenges to quantitative T1, T2 and diffusion imaging. Here, we present the MESMERISED sequence (Multiplexed Echo Shifted Multiband Excited and Recalled Imaging of STEAM Encoded Diffusion). MESMERISED removes the dead time in Stimulated Echo Acquisition Mode (STEAM) imaging by an echo-shifting mechanism. The echo-shift (ES) factor is independent of multiband (MB) acceleration and allows for very high multiplicative (ESxMB) acceleration factors, particularly under moderate and long mixing times. This results in high duty cycle and time efficiency at 7 T, particularly for quantitative T1 and diffusion imaging, while also retaining the capacity to perform quantitative T2 and B1 mapping. We demonstrate the super-acceleration of MESMERISED for whole-brain T1 relaxometry with total acceleration factors up to 36 at 1.8 mm isotropic resolution, and up to 54 at 1.25 mm resolution qT1 imaging, corresponding to a 6x and 9x speed-up, respectively, compared to MB-only accelerated acquisitions. We demonstrate quantitative T2 and B1 mapping to illustrate multicontrast mapping with the same MESMERISED sequence and the same readout train. Finally, we demonstrate highly efficient diffusion MRI with high b-values and long diffusion times in two separate cases. First, we show that super-accelerated multi-shell diffusion acquisitions with 370 whole-brain diffusion volumes over 8 b-value shells up to b = 7000 s/mm 2 can be generated at 2 mm isotropic in under 8 minutes, a data rate of almost a volume per second, or at 1.8 mm isotropic in under 11 minutes. Second, we demonstrate time-efficient sampling of different diffusion times with 1.8 mm isotropic diffusion data acquired at four diffusion times up to 290 ms, which supports both Diffusion Tensor Imaging (DTI) and Diffusion Kurtosis Imaging (DKI) at each diffusion time. MESMERISED extends possibilities to efficiently probe T1, T2 and diffusion contrast for multicomponent modeling of tissue microstructure.

show abstract

“…Low-rank denoising is a data-driven technique and does not incorporate prior knowledge or physical models of the data. These methods have also recently been applied to MR imaging techniques that utilise an additional dimension of encoding, such diffusion encoding direction in diffusion-weighted MRI (2,3), and time in functional MRI (4). Low-rank models have also been applied directly in the reconstruction of fast MRSI acquisitions (5,6).…”

Section: Introductionmentioning

confidence: 99%

Uncertainty in denoising of MRSI using low-rank methods

Clarke

Chiew

2021

Preprint

View full text Add to dashboard Cite

Purpose: Low-rank denoising of MRSI data results in an apparent increase in spectral SNR. However, it is not clear if this translates to a lower uncertainty in metabolite concentrations after spectroscopic fitting. Estimation of the true uncertainty after denoising is desirable for downstream analysis in spectroscopy. In this work the uncertainty reduction from low-rank denoising methods based on spatio-temporal separability and linear predictability in MRSI are assessed. A new method for estimating metabolite concentration uncertainty after denoising is proposed. Finally, automatic rank threshold selection methods are assessed in simulated low SNR regimes. Methods: Assessment of denoising methods is conducted using Monte Carlo simulation of proton MRSI data, and by reproducibility of repeated in vivo acquisitions in five subjects. Results: In simulated and in vivo data, spatio-temporal based denoising is shown to reduce the concentration uncertainty, but linear prediction denoising increases uncertainty. Uncertainty estimates provided by fitting algorithms after denoising consistently under-estimate actual metabolite uncertainty. However, the proposed uncertainty estimation, based on an analytical expression for entry-wise variance after denoising, is more accurate. Finally, it is shown automated rank threshold selection using Marchenko-Pastur distribution can bias the data in low SNR conditions. An alternative soft-thresholding function is proposed. Conclusion: Low-rank denoising methods based on spatio-temporal separability do reduce uncertainty in MRS(I) data. However, thorough assessment is needed as assessment by SNR measured from residual baseline noise is insufficient given the presence of non-uniform variance. It is also important to select the right rank thresholding method in low SNR cases.

show abstract

NOise reduction with DIstribution Corrected (NORDIC) PCA in dMRI with complex-valued parameter-free locally low-rank processing

Cited by 75 publications

References 35 publications

NORDIC Increases the Sensitivity and Preserves the Spatiotemporal Precision of fMRI Responses

NORDIC Increases the Sensitivity and Preserves the Spatiotemporal Precision of fMRI Responses

MESMERISED: Super-accelerating T₁relaxometry and diffusion MRI with STEAM at 7 T for quantitative multi-contrast and diffusion imaging

Uncertainty in denoising of MRSI using low-rank methods

Contact Info

Product

Resources

About

NOise reduction with DIstribution Corrected (NORDIC) PCA in dMRI with complex-valued parameter-free locally low-rank processing

Cited by 75 publications

References 35 publications

NORDIC Increases the Sensitivity and Preserves the Spatiotemporal Precision of fMRI Responses

NORDIC Increases the Sensitivity and Preserves the Spatiotemporal Precision of fMRI Responses

MESMERISED: Super-accelerating T1relaxometry and diffusion MRI with STEAM at 7 T for quantitative multi-contrast and diffusion imaging

Uncertainty in denoising of MRSI using low-rank methods

Contact Info

Product

Resources

About

MESMERISED: Super-accelerating T₁relaxometry and diffusion MRI with STEAM at 7 T for quantitative multi-contrast and diffusion imaging