Adaptive denoising for chemical exchange saturation transfer MR imaging

Breitling, Johannes; Deshmane, A; Goerke, Steffen; Korzowski, Andreas; Herz, K; Ladd, Mark E.; Scheffler, Klaus; Bachert, Peter; Zaiß, Moritz

doi:10.1002/nbm.4133

Cited by 38 publications

(73 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An overview of the presented data processing pipeline is given in Figure . In vivo Z‐spectra are acquired at 3T and evaluated by conventional methods (PCA denoising, Lorentzian fitting). The obtained results are used to train a deep neural network to directly map from raw Z‐spectra to Lorentzian parameters, yielding CEST contrast maps in ~1 s where conventional evaluation takes ~10 min.…”

Section: Methodsmentioning

confidence: 99%

“…Brain masks for each subject data set were created manually. The Z‐spectra were spectrally de‐noised by principal component analysis (PCA) following the approach of Breitling et al, keeping only the principal components suggested by the median criterion (Supporting Information Table ). To isolate CEST effects, a four‐pool Lorentzian fit model according to Goerke et al was used to fit the direct water saturation (DS), semi‐solid magnetization transfer (MT), amide proton transfer (APT), and relayed NOE peaks, using the model equation

Z (Δ ω) = c - L_{DS} - L_{APT} - L_{NOE} - L_{MT},

with a constant c, the direct saturation pool

L_{italicDS} = \frac{A_{DS}}{1 + {()(, \frac{Δ ω - {normalδ}_{DS}}{{normalΓ}_{DS} / 2})}^{2}},

and the other pools

L_{i} = \frac{A_{i}}{1 + {()(, \frac{Δ ω - {normalδ}_{DS} - {normalδ}_{i}}{{normalΓ}_{i} / 2})}^{2}}, i \in \{APT, NOE, MT\},

with amplitudes A i , full‐width‐at‐half‐maximum Γ i , and peak positions δ i .…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

DeepCEST 3T: Robust MRI parameter determination and uncertainty quantification with neural networks—application to CEST imaging of the human brain at 3T

Glang

Deshmane

Prokudin

et al. 2019

Magnetic Resonance in Med

Self Cite

View full text Add to dashboard Cite

Purpose Calculation of sophisticated MR contrasts often requires complex mathematical modeling. Data evaluation is computationally expensive, vulnerable to artifacts, and often sensitive to fit algorithm parameters. In this work, we investigate whether neural networks can provide not only fast model fitting results, but also a quality metric for the predicted values, so called uncertainty quantification, investigated here in the context of multi‐pool Lorentzian fitting of CEST MRI spectra at 3T. Methods A deep feed‐forward neural network including a probabilistic output layer allowing for uncertainty quantification was set up to take uncorrected CEST‐spectra as input and predict 3T Lorentzian parameters of a 4‐pool model (water, semisolid MT, amide CEST, NOE CEST), including the B0 inhomogeneity. Networks were trained on data from 3 subjects with and without data augmentation, and applied to untrained data from 1 additional subject and 1 brain tumor patient. Comparison to conventional Lorentzian fitting was performed on different perturbations of input data. Results The deepCEST 3T networks provided fast and accurate predictions of all Lorentzian parameters and were robust to input perturbations because of noise or B0 artifacts. The uncertainty quantification detected fluctuations in input data by increase of the uncertainty intervals. The method generalized to unseen brain tumor patient CEST data. Conclusions The deepCEST 3T neural network provides fast and robust estimation of CEST parameters, enabling online reconstruction of sophisticated CEST contrast images without the typical computational cost. Moreover, the uncertainty quantification indicates if the predictions are trustworthy, enabling confident interpretation of contrast changes.

show abstract

Section: Methodsmentioning

confidence: 99%

Z (Δ ω) = c - L_{DS} - L_{APT} - L_{NOE} - L_{MT},

with a constant c, the direct saturation pool

L_{italicDS} = \frac{A_{DS}}{1 + {()(, \frac{Δ ω - {normalδ}_{DS}}{{normalΓ}_{DS} / 2})}^{2}},

and the other pools

L_{i} = \frac{A_{i}}{1 + {()(, \frac{Δ ω - {normalδ}_{DS} - {normalδ}_{i}}{{normalΓ}_{i} / 2})}^{2}}, i \in \{APT, NOE, MT\},

with amplitudes A i , full‐width‐at‐half‐maximum Γ i , and peak positions δ i .…”

Section: Methodsmentioning

confidence: 99%

DeepCEST 3T: Robust MRI parameter determination and uncertainty quantification with neural networks—application to CEST imaging of the human brain at 3T

Glang

Deshmane

Prokudin

et al. 2019

Magnetic Resonance in Med

Self Cite

View full text Add to dashboard Cite

show abstract

“…The resulting Z ‐data were first denoised using an algorithm based on principal component analysis with the number of components determined according to the Malinowski criterion, and subsequently averaged to obtain the required nine Z ‐images. The final T protein contrast—as calculated from Equations and —was corrected for B 1 inhomogeneities by the “one‐point‐contrast correction” method .…”

Section: Methodsmentioning

confidence: 99%

“…To achieve this goal we applied several methodological developments: (i) implementation of cosine‐modulated pulses allowed for a simultaneous presaturation at two frequency offsets; and was additionally combined with (ii) a weighted acquisition scheme for optimal averaging. Furthermore, (iii) utilization of a 3D snapshot‐CEST image readout allowed for retrospective motion correction and (iv) facilitated the application of advanced denoising strategies . Preservation of specificity of the optimized dualCEST protocol was demonstrated in vitro.…”

Section: Introductionmentioning

confidence: 99%

Optimized dualCEST‐MRI for imaging of endogenous bulk mobile proteins in the human brain

et al. 2020

Self Cite

View full text Add to dashboard Cite

Dual-frequency irradiation chemical exchange saturation transfer (dualCEST) allows imaging of endogenous bulk mobile proteins by selectively measuring the intramolecular spin diffusion. The resulting specificity to changes in the concentration, molecular size, and folding state of mobile proteins is of particular interest as a marker for neurodegenerative diseases and cancer. Until now, application of dualCEST in clinical trials was prevented by the inherently small signal-to-noise ratio and the resulting comparatively long examination time. In this study, we present an optimized acquisition protocol allowing 3D dualCEST-MRI examinations in a clinically relevant time frame. The optimization comprised the extension of the image readout to 3D, allowing a retrospective co-registration and application of denoising strategies. In addition, cosine-modulated dual-frequency presaturation pulses were implemented with a weighted acquisition scheme of the necessary frequency offsets. The optimization resulted in a signal-to-noise ratio gain by a factor of approximately 8. In particular, the application of denoising and the motion correction were the most crucial improvement steps. In vitro experiments verified the preservation of specificity of the dualCEST signal to proteins. Good-to-excellent intra-session and good intersession repeatability was achieved, allowing reliable detection of relative signal differences of about 16% or higher. Applicability in a clinical setting was demonstrated by examining a patient with glioblastoma. The optimized acquisition protocol for dualCEST-MRI at 3 T enables selective imaging of endogenous bulk mobile proteins under clinically relevant conditions. K E Y W O R D S

show abstract

“…A joint authorship statement was mistakenly omitted from the article by Breitling et al The statement is given here:…”

mentioning

confidence: 99%

Erratum

2020

NMR in Biomedicine

View full text Add to dashboard Cite

Adaptive denoising for chemical exchange saturation transfer MR imaging

Cited by 38 publications

References 56 publications

DeepCEST 3T: Robust MRI parameter determination and uncertainty quantification with neural networks—application to CEST imaging of the human brain at 3T

DeepCEST 3T: Robust MRI parameter determination and uncertainty quantification with neural networks—application to CEST imaging of the human brain at 3T

Optimized dualCEST‐MRI for imaging of endogenous bulk mobile proteins in the human brain

Erratum

Contact Info

Product

Resources

About