Acquiring and Predicting Multidimensional Diffusion (MUDI) Data: An Open Challenge

Pizzolato, Marco; Palombo, Marco; Bonet‐Carne, Elisenda; Tax, Chantal M. W.; Grussu, Francesco; Ianuş, Andrada; Bogusz, Fabian; Pieciak, Tomasz; Ning, Lipeng; Larochelle, Hugo; Descoteaux, Maxime; Chamberland, Maxime; Blumberg, Stefano B.; Mertzanidou, Thomy; Alexander, Daniel C.; Afzali, Maryam; Aja‐Fernández, Santiago; Jones, Derek K.; Westin, Carl‐Fredrik; Rathi, Yogesh; Baete, Steven H.; Cordero‐Grande, Lucilio; Ladner, Thilo; Slator, Paddy J.; Hajnal, Joseph V.; Thiran, Jean‐Philippe; Price, Anthony N.; Sepehrband, Farshid; Zhang, Fan; Hutter, Jana

doi:10.1007/978-3-030-52893-5_17

Cited by 16 publications

(29 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this work, we tested the learning-based approach on protocols designed via Cramer-Rao lower-bound optimization (Alexander, 2008; Coelho et al, 2019; Lampinen et al, 2020). While not yet tested in conjunction with learning-based fitting pipelines, alternative optimization methods based on either efficient signal decomposition schemes (Bates et al, 2020; Song and Xiao, 2020) or deep learning algorithms for feature selection (Grussu et al, 2020b; Pizzolato et al, 2020) are also expected to have a positive impact on the performance of the DNN fitting approach.…”

Section: Discussionmentioning

confidence: 99%

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

Martins

Nilsson

Lampinen

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Specific features of white-matter microstructure can be investigated by using biophysical models to interpret relaxation-diffusion MRI brain data. Although more intricate models have the potential to reveal more details of the tissue, they also incur time-consuming parameter estimation that may converge to inaccurate solutions due to a prevalence of local minima in a degenerate fitting landscape. Machine-learning fitting algorithms have been proposed to accelerate the parameter estimation and increase the robustness of the attained estimates. So far, learning-based fitting approaches have been restricted to lower-dimensional microstructural models where dense sets of training data are easy to generate. Moreover, the degree to which machine learning can alleviate the degeneracy problem is poorly understood. For conventional least-squares solvers, it has been shown that degeneracy can be avoided by acquisition with optimized relaxation-diffusion-correlation protocols that include tensor-valued diffusion encoding; whether machine-learning techniques can offset these acquisition requirements remains to be tested. In this work, we employ deep neural networks to vastly accelerate the fitting of a recently introduced high-dimensional relaxation-diffusion model of tissue microstructure. We also develop strategies for assessing the accuracy and sensitivity of function fitting networks and use those strategies to explore the impact of acquisition protocol design on the performance of the network. The developed learning-based fitting pipelines were tested on relaxation-diffusion data acquired with optimized and sub-sampled acquisition protocols. We found no evidence that machine-learning algorithms can by themselves replace a careful design of the acquisition protocol or correct for a degenerate fitting landscape.

show abstract

Section: Discussionmentioning

confidence: 99%

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

Martins

Nilsson

Lampinen

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Open challenges play an important role to gain a better understanding of how various models capture the dMRI signal decay, as they put forward rich datasets and well-defined tasks and usually receive submissions from across the modelling landscapes (Uran Ferizi et al 2017; Schilling et al 2019; Pizzolato et al 2020). The last diffusion microstructure challenge which included a comprehensive dMRI acquisition (Uran Ferizi et al 2017) was organized in 2015 and focused on modelling the dMRI signal acquired on the Connectome scanner for two ROIs in white matter: genu of the Corpus Callosum with mostly aligned fibres and fornix with a more complex fibre configuration.…”

Section: Introductionmentioning

confidence: 99%

“…Since the end of this challenge, many novel approaches have been proposed, including a booming application of machine learning techniques for data fitting and prediction (Golkov et al 2016; Nedjati-Gilani et al 2017; Nath, Schilling, et al 2019; Ravi et al 2019; Poulin et al 2019). Moreover, previous challenges (Uran Ferizi et al 2017; Schilling et al 2019; Pizzolato et al 2020) included only diffusion data acquired with standard SDE sequences, and do not provide any insight into the different approaches available to analyse advanced sequences such as DDE.…”

Section: Introductionmentioning

confidence: 99%

“…In this challenge we set to evaluate the ability of different dMRI modeling approaches to capture the dMRI signal contrast from state-of-the-art acquisitions performed with SDE, including shells with an order of magnitude higher angular resolution compared to previous challenges (Uran Ferizi et al 2017;Schilling et al 2019;Pizzolato et al 2020), as well as DDE and DODE data. Further, we aim to investigate the relationship between the goodness of fit and tissue type, acquisition parameters, and diffusion sensitization.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

On the generalizability of diffusion MRI signal representations across acquisition parameters, sequences and tissue types: chronicles of the MEMENTO challenge

Luca

Ianuş

Leemans

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Diffusion MRI (dMRI) has become an invaluable tool to assess the microstructural organization of brain tissue. Depending on the specific acquisition settings, the dMRI signal encodes specific properties of the underlying diffusion process. In the last two decades, several signal representations have been proposed to fit the dMRI signal and decode such properties. Most methods, however, are tested and developed on a limited amount of data, and their applicability to other acquisition schemes remains unknown. With this work, we aimed to shed light on the generalizability of existing dMRI signal representations to different diffusion encoding parameters and brain tissue types. To this end, we organized a community challenge - named MEMENTO, making available the same datasets for fair comparisons across algorithms and techniques. We considered two state-of-the-art diffusion datasets, including single-diffusion-encoding (SDE) spin-echo data from a human brain with over 3820 unique diffusion weightings (the MASSIVE dataset), and double (oscillating) diffusion encoding data (DDE/DODE) of a mouse brain including over 2520 unique data points. A subset of the data sampled in 5 different voxels was openly distributed, and the challenge participants were asked to predict the remaining part of the data. After one year, eight participant teams submitted a total of 80 signal fits. For each submission, we evaluated the mean squared error, the variance of the prediction error and the Bayesian information criteria. Most predictions predicted either multi-shell SDE data (37%) or DODE data (22%), followed by cartesian SDE data (19%) and DDE (18%). Most submissions predicted the signals measured with SDE remarkably well, with the exception of low and very strong diffusion weightings. The prediction of DDE and DODE data seemed more challenging, likely because none of the submissions explicitly accounted for diffusion time and frequency. Next to the choice of the model, decisions on fit procedure and hyperparameters play a major role in the prediction performance, highlighting the importance of optimizing and reporting such choices. This work is a community effort to highlight strength and limitations of the field at representing dMRI acquired with trending encoding schemes, gaining insights into how different models generalize to different tissue types and fiber configurations over a large range of diffusion encodings.

show abstract

“…In the ZEBRA approach, a relatively long TR is used to acquire additional contrasts for multi-parametric characterization of the tissue. 29,30 For example, ZEBRA includes multiple gradient echoes after the first diffusion-encoded spin-echo to map the T * 2 time. There is a trade-off between the number of TEs and TIs sampled within a TR, which will impact the resulting T * 2 and T 1 maps, respectively.…”

Section: Introductionmentioning

confidence: 99%

Efficient whole‐brain tract‐specific T₁ mapping at 3T with slice‐shuffled inversion‐recovery diffusion‐weighted imaging

Leppert

Andrews

Campbell

et al. 2021

Magnetic Resonance in Med

View full text Add to dashboard Cite

Most voxels in white matter contain multiple fiber populations with different orientations and levels of myelination. Conventional T 1 mapping measures 1 T 1 value per voxel, representing a weighted average of the multiple tract T 1 times.Inversion-recovery diffusion-weighted imaging (IR-DWI) allows the T 1 times of multiple tracts in a voxel to be disentangled, but the scan time is prohibitively long.Recently, slice-shuffled IR-DWI implementations have been proposed to significantly reduce scan time. In this work, we demonstrate that we can measure tractspecific T 1 values in the whole brain using simultaneous multi-slice slice-shuffled IR-DWI at 3T. Methods: We perform simulations to evaluate the accuracy and precision of our crossing fiber IR-DWI signal model for various fiber parameters. The proposed sequence and signal model are tested in a phantom consisting of crossing asparagus pieces doped with gadolinium to vary T 1 , and in 2 human subjects. Results: Our simulations show that tract-specific T 1 times can be estimated within 5% of the nominal fiber T 1 values. Tract-specific T 1 values were resolved in subvoxel 2 fiber crossings in the asparagus phantom. Tract-specific T 1 times were resolved in 2 different tract crossings in the human brain where myelination differences have previously been reported; the crossing of the cingulum and genu of the corpus callosum and the crossing of the corticospinal tract and pontine fibers. Conclusion: Whole-brain tract-specific T 1 mapping is feasible using slice-shuffled IR-DWI at 3T. This technique has the potential to improve the microstructural How to cite this article: Leppert IR, Andrews DA, Campbell JSW, et al. Efficient whole-brain tract-specific T 1 mapping at 3T with slice-shuffled inversion-recovery diffusion-weighted imaging.

show abstract

Acquiring and Predicting Multidimensional Diffusion (MUDI) Data: An Open Challenge

Cited by 16 publications

References 15 publications

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

On the generalizability of diffusion MRI signal representations across acquisition parameters, sequences and tissue types: chronicles of the MEMENTO challenge

Efficient whole‐brain tract‐specific T₁ mapping at 3T with slice‐shuffled inversion‐recovery diffusion‐weighted imaging

Contact Info

Product

Resources

About

Acquiring and Predicting Multidimensional Diffusion (MUDI) Data: An Open Challenge

Cited by 16 publications

References 15 publications

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

Neural Networks for parameter estimation in microstructural MRI: a study with a high-dimensional diffusion-relaxation model of white matter microstructure

On the generalizability of diffusion MRI signal representations across acquisition parameters, sequences and tissue types: chronicles of the MEMENTO challenge

Efficient whole‐brain tract‐specific T1 mapping at 3T with slice‐shuffled inversion‐recovery diffusion‐weighted imaging

Contact Info

Product

Resources

About

Efficient whole‐brain tract‐specific T₁ mapping at 3T with slice‐shuffled inversion‐recovery diffusion‐weighted imaging