Myelin water imaging data analysis in less than one minute

Liu, Hanwen; Xiang, Qing; Tam, Roger; Dvorak, Adam; MacKay, Alex L.; Kolind, Shannon; Traboulsee, Anthony; Vavasour, Irene M.; Li, David K.B.; Kramer, J. K. G.; Laule, Cornelia

doi:10.1016/j.neuroimage.2020.116551

Cited by 33 publications

(48 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Each hidden layer consists of standard linear operators followed by rectified non-linear units (ReLU(x) = max(0, x)). The number of hidden neurons decreases linearly from N to P , but other architectures are also possible [4,7]. The DNN is trained with standard back-propagation on synthetic or in vivo MRI data.…”

Section: Qmri Model Fitting With Dnnsmentioning

confidence: 99%

“…Alternative definitions of f are also in use to account for non-Gaussian noise or in maximum a posteriori estimation [3]. Finally, f can also include regularisers to stabilise the forward model inversion, as for example terms proportional to p 2 or p 1 (Tikhonov or l 1 regularisation) [4].…”

Section: Introductionmentioning

confidence: 99%

“…Machine learning has been proposed as an alternative to conventional NLLS to perform qMRI model fitting in qMRI and overcome its limitations. Notable examples include permeability [6] and soma [8] diffusion-weighted (DW) imaging via random forest regression trees, or q-space DW imaging [9] and myelin mapping [4] with deep neural networks (DNNs). Machine learning approaches can be trained off-line and then deployed almost instantaneously when new data come in [4], and offer more stable solutions when trained accurately [6,7], given their excellent function approximation properties [10].…”

Section: Introductionmentioning

confidence: 99%

“…In this work we introduce an implementation of qMRI fitting based on fully-connected DNNs and use it to compare systematically different training strategies, since at present it is not known which strategy leads to the most accurate, precise and robust parameter estimation. Specifically, we trained DNNs by optimising loss functions defined either i) as the l 2 -norm of tissue parameter estimation errors [6,4,8], or ii) the l 2 -norm of MRI signal estimation errors [7]. We compared the two learning approaches using saturation inversion recovery (SIR) DW data obtained at 3T for joint diffusion and T1 mapping, and performed computational experiments both in silico and in vivo.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Deep learning model fitting for diffusion-relaxometry: a comparative study

Grussu¹,

Battiston²,

Palombo³

et al. 2020

Preprint

View full text Add to dashboard Cite

Quantitative Magnetic Resonance Imaging (qMRI) signal model fitting is traditionally performed via non-linear least square (NLLS) estimation. NLLS is slow and its performance can be affected by the presence of different local minima in the fitting objective function. Recently, machine learning techniques, including deep neural networks (DNNs), have been proposed as robust alternatives to NLLS. Here we present a deep learning implementation of qMRI model fitting, which uses DNNs to perform the inversion of the forward signal model. We compare two DNN training strategies, based on two alternative definitions of the loss function, since at present it is not known which definition leads to the most accurate, precise and robust parameter estimation. In strategy 1 we define the loss as the l2-norm of tissue parameter prediction errors, while in strategy 2 as the l2-norm of MRI signal prediction errors. We compare the two approaches on synthetic and 3T in vivo saturation inversion recovery (SIR) diffusion-weighted (DW) MRI data, using a model for joint diffusion-T1 mapping. Strategy 1 leads to lower tissue parameter root mean squared errors (RMSEs) when realistic noise distributions are considered (e.g. Rician vs Gaussian). However, strategy 2 offers lower signal reconstruction RMSE, and allows training to be performed on both synthetic and actual in vivo MRI measurements. In conclusion, both strategies are valid choices for DNN-based fitting. Strategy 2 is more practical, as it does not require pre-computation of reference tissue parameters, but may lead to worse parameter estimation.

show abstract

Section: Qmri Model Fitting With Dnnsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Deep learning model fitting for diffusion-relaxometry: a comparative study

Grussu¹,

Battiston²,

Palombo³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…in the brain [4] and to estimate T1 and T2 in a fast and robust way using sparse data from magnetic resonance fingerprinting [5]; whereas ML methods based on convolutional neural network approaches have been developed to estimate susceptibility using a single subject orientation [6]. In dMRI, ML has been used, for example, to bridge the gap between datahungry imaging techniques and clinically feasible scans, for example by reconstructing superresolved maps from low spatial resolution data [7] [8], or by estimating advanced diffusionbased metrics from sparse q-space acquisitions [9] [10] [11].…”

mentioning

confidence: 99%

Training Data Distribution Significantly Impacts the Estimation of Tissue Microstructure with Machine Learning

Gyori

Palombo

Clark

et al. 2021

Preprint

View full text Add to dashboard Cite

Purpose: Supervised machine learning (ML) provides a compelling alternative to traditional model fitting for parameter mapping in quantitative MRI. The aim of this work is to demonstrate and quantify the effect of different training strategies on the accuracy and precision of parameter estimates when supervised ML is used for fitting. Methods: We fit a two-compartment biophysical model to diffusion measurements from in-vivo human brain, as well as simulated diffusion data, using both traditional model fitting and supervised ML. For supervised ML, we train several artificial neural networks, as well as random forest regressors, on different distributions of ground truth parameters. We compare the accuracy and precision of parameter estimates obtained from the different estimation approaches using synthetic test data. Results: When the distribution of parameter combinations in the training set matches those observed in similar data sets, we observe high precision, but inaccurate estimates for atypical parameter combinations. In contrast, when training data is sampled uniformly from the entire plausible parameter space, estimates tend to be more accurate for atypical parameter combinations but may have lower precision for typical parameter combinations. Conclusion: This work highlights the need to consider the choice of training data when deploying supervised ML for estimating microstructural metrics, as performance depends strongly on the training-set distribution. We show that high precision obtained using ML may mask strong bias, and visual assessment of the parameter maps is not sufficient for evaluating the quality of the estimates.

show abstract

An Unsupervised Deep Learning Approach for Dynamic‐Exponential Intravoxel Incoherent Motion MRI Modeling and Parameter Estimation in the Liver

Zhou

Wang

Liu

et al. 2022

Magnetic Resonance Imaging

View full text Add to dashboard Cite

BackgroundDynamic‐exponential intravoxel incoherent motion (IVIM) imaging is a potential technique for prediction, monitoring, and differential diagnosis of hepatic diseases, especially liver tumors. However, the use of such technique at voxel level is still limited.PurposeTo develop an unsupervised deep learning approach for voxel‐wise dynamic‐exponential IVIM modeling and parameter estimation in the liver.Study TypeProspective.PopulationTen healthy subjects (4 males; age 28 ± 6 years).Field Strength/SequenceSingle‐shot spin‐echo echo planar imaging (SE‐EPI) sequence with monopolar diffusion‐encoding gradients (12 b‐values, 0–800 seconds/mm2) at 3.0 T.AssessmentThe proposed deep neural network (DNN) was separately trained on simulated and in vivo hepatic IVIM datasets. The trained networks were compared to the approach combining least squares with Akaike information criterion (LSQ‐AIC) in terms of dynamic‐exponential modeling accuracy, inter‐subject coefficients of variation (CVs), and fitting residuals on the simulated subsets and regions of interest (ROIs) in the left and right liver lobes. The ROIs were delineated by a radiologist (H.‐X.Z.) with 7 years of experience in MRI reading.Statistical TestsComparisons between approaches were performed with a paired t‐test (normality) or a Wilcoxon rank‐sum test (nonnormality). P < 0.05 was considered statistically significant.ResultsIn simulations, DNN gave significantly higher accuracy (91.6%–95.5%) for identification of bi‐exponential decays with respect to LSQ‐AIC (79.7%–86.8%). For tri‐exponential identification, DNN was also superior to LSQ‐AIC despite not reaching a significant level (P = 0.08). Additionally, DNN always yielded comparatively low root‐mean‐square error for estimated parameters. For the in vivo IVIM measurements, inter‐subject CVs (0.011–0.150) of DNN were significantly smaller than those (0.049–0.573) of LSQ‐AIC. Concerning fitting residuals, there was no significant difference between the two approaches (P = 0.56 and 0.76) in both the simulated and in vivo studies.Data ConclusionThe proposed DNN is recommended for accurate and robust dynamic‐exponential modeling and parameter estimation in hepatic IVIM imaging.Level of Evidence2Technical EfficacyStage 1

show abstract

Myelin water imaging data analysis in less than one minute

Cited by 33 publications

References 44 publications

Deep learning model fitting for diffusion-relaxometry: a comparative study

Deep learning model fitting for diffusion-relaxometry: a comparative study

Training Data Distribution Significantly Impacts the Estimation of Tissue Microstructure with Machine Learning

An Unsupervised Deep Learning Approach for Dynamic‐Exponential Intravoxel Incoherent Motion MRI Modeling and Parameter Estimation in the Liver

Contact Info

Product

Resources

About