Simulation-based evaluation of susceptibility distortion correction methods in diffusion MRI for connectivity analysis

Estéban, Oscar; Daducci, Alessandro; Caruyer, Emmanuel; O’Brien, Kieran; Ledesma-Carbayo, María J.; Cuadra, M. Bach; Santos, A.

doi:10.1109/isbi.2014.6867976

Cited by 11 publications

(27 citation statements)

References 12 publications

(13 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, the reader may wish to compare the relative performance of different distortion compensation methods in their own dataset, using an approach similar to ours. We used multiple metrics to quantify EPI-T1 agreement as a proxy for correction quality (i.e., Dice coefficients for whole-brain masks and CSF-excluded masks, as well as mutual information), since we acknowledge that there is no single gold standard for measuring the quality of distortion compensation in human brain imaging data 28 (but see the following studies that used simulations to try to establish ground truth 47,48 ). By making our data and analysis code publicly available (see Methods), we hope to facilitate the empirical selection of effective approaches for geometric distortion compensation in future research.…”

Section: Discussionmentioning

confidence: 99%

“…In particular, previous work has generally shown better performance for oppPE field map strategies, as compared to B0 field maps, which has been attributed in part to the difficulty of using B0 field maps to correct distortion near the edges of the brain (see Figure 1F), where phase values change rapidly. Using simulated EPI data, both Esteban 47 and Graham 48 showed quantitatively that ground truth undistorted images were recovered best using an oppPE field map method, whereas B0 field maps performed slightly worse, and nonlinear registration-based methods were greatly inferior (but still better than no correction at all). Similar conclusions were reached by Hong and colleagues 28 using SE EPI in the mouse brain at 7T, by Holland and colleagues 34 using SE EPI at 1.5 and 3T in the human brain, and by Wang and colleagues 49 using 3T dMRI data in humans (see also 5 ).…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Assessing methods for geometric distortion compensation in 7T gradient echo fMRI data

Schallmo

Weldon

Sponheim

et al. 2020

Preprint

View full text Add to dashboard Cite

AbstractEcho planar imaging (EPI) is widely used in functional and diffusion-weighted MRI, but suffers from significant geometric distortions in the phase encoding direction caused by inhomogeneities in the static magnetic field (B0). This is a particular challenge for EPI at very high field (7T and above), as distortion increases with higher field strength. A number of techniques for correcting geometric distortion exist, including those based on B0 field mapping and acquiring EPI scans with opposite phase encoding directions. However, few quantitative comparisons of distortion compensation methods have been performed using EPI data from the human brain, and even fewer at very high field. In the current study, we compared geometric distortion compensation using B0 field maps and opposite phase encoding scans implemented in two different software packages (FSL and AFNI) applied to 7T gradient echo EPI data from 31 human participants. We assessed the quality of distortion compensation by quantifying the degree of alignment to a T1-weighted anatomical reference scan using Dice coefficients and mutual information. We found that the best distortion compensation was achieved in our dataset using gradient echo scans with opposite phase encoding directions to map the distortion, as compared to B0 field maps or spin echo opposite phase encoding scans. Performance between FSL and AFNI was equivalent. While the ideal geometric distortion compensation approach may vary due to methodological differences across experiments, this study provides a framework for researchers to assess the quality of different distortion compensation methods in their own work.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Assessing methods for geometric distortion compensation in 7T gradient echo fMRI data

Schallmo

Weldon

Sponheim

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…For example, it is common to collect LR (left right) and RL (right left) images, or AP (anterior posterior) and PA (posterior anterior) images. Phase encoding based techniques have been demonstrated to outperform the other two approaches (Graham et al, 2017;Esteban et al, 2014), at the cost of a longer scan time.…”

Section: Introductionmentioning

confidence: 99%

“…The lack of ground truth means that evaluations are typically indirect or qualitative (Jezzard and Balaban, 1995;Wu et al, 2008;Bhushan et al, 2012;Ruthotto et al, 2013;Fritz et al, 2014;Irfanoglu et al, 2015;Taylor et al, 2016;Hedouin et al, 2017;Wang et al, 2017;Irfanoglu et al, 2018). Only a few investigations have been carried out with the presence of a ground truth for evaluation of susceptibility distortion correction (Andersson et al, 2003;Esteban et al, 2014;Graham et al, 2017). Hedouin et al (2017) compared aDC, aBMDC and TOPUP using phantom data and human data.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of six phase encoding based susceptibility distortion correction methods for diffusion MRI

Eklund

2019

Preprint

View full text Add to dashboard Cite

ITEA3 / VINNOVA funded project "Intelligence based iMprovement of Personalized treatment And Clinical workflow supporT" (IMPACT) Purpose: Susceptibility distortion impacts diffusion MRI data analysis and is typically corrected during preprocessing. Correction strategies involve three classes of methods: registration to a structural image, the use of a fieldmap or the use of images acquired with opposing phase encoding directions. It has been demonstrated that phase encoding based methods outperform the other two classes but unfortunately, the choice of which phase encoding based method to use is still an open question due to the absence of any systematic comparisons. Methods:In this paper we quantitatively evaluated six popular phase encoding based methods for correcting susceptibility distortions in diffusion MRI data. We employed a framework that allows for the simulation of realistic diffusion MRI data with susceptibility distortions. We evaluated the ability for methods to correct distortions by comparing the corrected data with the ground truth. Four diffusion tensor metrics (FA, MD, eigenvalues and eigenvectors) were calculated from the corrected data and compared with the ground truth. We also validated two popular indirect metrics using both simulated data and real data. The two indirect metrics are the difference between the corrected LR and AP data, and the FA standard deviation over the corrected LR, RL, AP and PA data. 1 2 XUAN GU ET AL. Results:We found that TOPUP offered the most accurate and robust correction than the other five methods when evaluating using both direct and indirect metrics. EPIC and HySCO performed well in correcting b 0 images but produced poor corrections for diffusion weighted volumes and they produced large errors for the four diffusion tensor metrics.We also demonstrated that the indirect metric gave a different ordering of correction quality than the direct metric. Conclusion:We suggest researchers to use TOPUP for susceptibility distortion correction. We also suggest that indirect metrics must be interpreted cautiously when evaluating methods for correcting susceptibility distortions in diffusion MRI data.

show abstract

“…Diffantom is originally designed for the investigation of susceptibility-derived distortions, a typical artifact that produces geometrical warping in certain regions of dMRI datasets. In Esteban et al (2014) we addressed this phenomenon and concluded that the connectivity matrix of 1 Available at: http://www.tractometer.org/ismrm_2015_challenge/.…”

Section: Introductionmentioning

confidence: 99%

Diffantom: whole-brain diffusion MRI phantoms derived from real datasets of the Human Connectome Project

Estéban

Caruyer

Daducci

et al. 2015

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Simulation-based evaluation of susceptibility distortion correction methods in diffusion MRI for connectivity analysis

Cited by 11 publications

References 12 publications

Assessing methods for geometric distortion compensation in 7T gradient echo fMRI data

Assessing methods for geometric distortion compensation in 7T gradient echo fMRI data

Evaluation of six phase encoding based susceptibility distortion correction methods for diffusion MRI

Diffantom: whole-brain diffusion MRI phantoms derived from real datasets of the Human Connectome Project

Contact Info

Product

Resources

About