A supervised deep neural network approach with standardized targets for enhanced accuracy of IVIM parameter estimation from multi‐SNR images

PurposeJoint analysis of flow‐compensated (FC) and non‐flow‐compensated (NC) diffusion MRI (dMRI) data has been suggested for increased robustness of intravoxel incoherent motion (IVIM) parameter estimation. For this purpose, a set of methods commonly used or previously found useful for IVIM analysis of dMRI data obtained with conventional diffusion encoding were evaluated in healthy human brain.MethodsFive methods for joint IVIM analysis of FC and NC dMRI data were compared: (1) direct non‐linear least squares fitting, (2) a segmented fitting algorithm with estimation of the diffusion coefficient from higher b‐values of NC data, (3) a Bayesian algorithm with uniform prior distributions, (4) a Bayesian algorithm with spatial prior distributions, and (5) a deep learning‐based algorithm. Methods were evaluated on brain dMRI data from healthy subjects and simulated data at multiple noise levels. Bipolar diffusion encoding gradients were used with b‐values 0–200 s/mm2 and corresponding flow weighting factors 0–2.35 s/mm for NC data and by design 0 for FC data. Data were acquired twice for repeatability analysis.ResultsMeasurement repeatability as well as estimation bias and variability were at similar levels or better with the Bayesian algorithm with spatial prior distributions and the deep learning‐based algorithm for IVIM parameters and , and for the Bayesian algorithm only for , relative to the other methods.ConclusionA Bayesian algorithm with spatial prior distributions is preferable for joint IVIM analysis of FC and NC dMRI data in the healthy human brain, but deep learning‐based algorithms appear promising.

Section: Discussionmentioning

confidence: 99%

Comparison of methods for intravoxel incoherent motion parameter estimation in the brain from flow‐compensated and non‐flow‐compensated diffusion‐encoded data

Jalnefjord,

Björkman‐Burtscher

2024

Section: Discussionmentioning

confidence: 99%

“…Moreover, supervised neural networks have been proposed as IVIM fitting approach, which are trained on simulated signal decay curves with matching IVIM parameter labels 27,28 . Similar to the unsupervised PINN, these are robust to image noise and allow clear separation between different tissue types 27,28 . However, the application of supervised neural networks to in vivo data is more challenging than PINNs due to the possible mismatch between the in vivo data and the synthesized data used for training 29 …”

Section: Discussionmentioning

confidence: 99%

“…27,28 Similar to the unsupervised PINN, these are robust to image noise and allow clear separation between different tissue types. 27,28 However, the application of supervised neural networks to in vivo data is more challenging than PINNs due to the possible mismatch between the in vivo data and the synthesized data used for training. 29 Regarding the unsupervised PINN, recent studies started employing more advanced network architectures and learning strategies.…”

Section: Other Ivim Model Fitting Approachesmentioning

confidence: 99%

See 1 more Smart Citation

Improving microstructural integrity, interstitial fluid, and blood microcirculation images from multi‐b‐value diffusion MRI using physics‐informed neural networks in cerebrovascular disease

Voorter

Backes

Gurney‐Champion

et al. 2023

Purpose To obtain better microstructural integrity, interstitial fluid, and microvascular images from multi‐b‐value diffusion MRI data by using a physics‐informed neural network (PINN) fitting approach. Methods Test–retest whole‐brain inversion recovery diffusion‐weighted images with multiple b‐values (IVIM: intravoxel incoherent motion) were acquired on separate days for 16 patients with cerebrovascular disease on a 3.0T MRI system. The performance of the PINN three‐component IVIM (3C‐IVIM) model fitting approach was compared with conventional fitting approaches (i.e., non‐negative least squares and two‐step least squares) in terms of (1) parameter map quality, (2) test–retest repeatability, and (3) voxel‐wise accuracy. Using the in vivo data, the parameter map quality was assessed by the parameter contrast‐to‐noise ratio (PCNR) between normal‐appearing white matter and white matter hyperintensities, and test–retest repeatability was expressed by the coefficient of variation (CV) and intraclass correlation coefficient (ICC). The voxel‐wise accuracy of the 3C‐IVIM parameters was determined by 10,000 computer simulations mimicking our in vivo data. Differences in PCNR and CV values obtained with the PINN approach versus conventional fitting approaches were assessed using paired Wilcoxon signed‐rank tests. Results The PINN‐derived 3C‐IVIM parameter maps were of higher quality and more repeatable than those of conventional fitting approaches, while also achieving higher voxel‐wise accuracy. Conclusion Physics‐informed neural networks enable robust voxel‐wise estimation of three diffusion components from the diffusion‐weighted signal. The repeatable and high‐quality biological parameter maps generated with PINNs allow for visual evaluation of pathophysiological processes in cerebrovascular disease.

“…[14][15][16] Other approaches address MRI signal denoising before the tissue parameter reconstruction, for example, with Marchenko-Pastur Principal Component Analysis 17 or Beltrami Denoising. [18][19][20] More modern data-driven deep learning approaches include mapping relaxation parameters with residual networks, 21,22 frameworks with physical model constraints, [23][24][25][26] supervised 27 and unsupervised 28 intravoxel incoherent motion estimation and models that address uncertainty estimation in dynamic contrast-enhanced-MRI 29 and ADC mapping. 30 To some degree, any method that incorporates denoising will rely on specific prior knowledge rather than the data.…”

Section: Introductionmentioning

confidence: 99%

Denoising and uncertainty estimation in parameter mapping with approximate Bayesian deep image priors

Hellström

Löfstedt

Garpebring

2023

PurposeTo mitigate the problem of noisy parameter maps with high uncertainties by casting parameter mapping as a denoising task based on Deep Image Priors.MethodsWe extend the concept of denoising with Deep Image Prior (DIP) into parameter mapping by treating the output of an image‐generating network as a parametrization of tissue parameter maps. The method implicitly denoises the parameter mapping process by filtering low‐level image features with an untrained convolutional neural network (CNN). Our implementation includes uncertainty estimation from Bernoulli approximate variational inference, implemented with MC dropout, which provides model uncertainty in each voxel of the denoised parameter maps. The method is modular, so the specifics of different applications (e.g., T1 mapping) separate into application‐specific signal equation blocks. We evaluate the method on variable flip angle T1 mapping, multi‐echo T2 mapping, and apparent diffusion coefficient mapping.ResultsWe found that deep image prior adapts successfully to several applications in parameter mapping. In all evaluations, the method produces noise‐reduced parameter maps with decreased uncertainty compared to conventional methods. The downsides of the proposed method are the long computational time and the introduction of some bias from the denoising prior.ConclusionDIP successfully denoise the parameter mapping process and applies to several applications with limited hyperparameter tuning. Further, it is easy to implement since DIP methods do not use network training data. Although time‐consuming, uncertainty information from MC dropout makes the method more robust and provides useful information when properly calibrated.