• Corpus ID: 212743984

Overview of Scanner Invariant Representations

  title={Overview of Scanner Invariant Representations},
  author={Daniel Moyer and Greg Ver Steeg and Paul M. Thompson},
Pooled imaging data from multiple sources is subject to bias from each source. Studies that do not correct for these scanner/site biases at best lose statistical power, and at worst leave spurious correlations in their data. Estimation of the bias effects is non-trivial due to the paucity of data with correspondence across sites, so called "traveling phantom" data, which is expensive to collect. Nevertheless, numerous solutions leveraging direct correspondence have been proposed. In contrast to… 


Bayesian Image Quality Transfer with CNNs: Exploring Uncertainty in dMRI Super-Resolution
This work proposes to account for intrinsic uncertainty through a per-patch heteroscedastic noise model and for parameter uncertainty through approximate Bayesian inference in the form of variational dropout and shows that the combined benefits lead to the state-of-the-art performance SR of diffusion MR brain images in terms of errors compared to ground truth.
Exploration of scanning effects in multi-site structural MRI studies
Both GLM and SBM correction can reduce scanning effects, though SBM's data-driven nature provides additional flexibility and is better able to handle collinear effects.
Multi-site harmonization of diffusion MRI data in a registration framework
A novel approach to harmonize dMRI data using rotation invariant spherical harmonic (RISH) features embedded within a multi-modal image registration framework and demonstrates that any abnormality in diffusion measures due to disease is preserved during the harmonization process.
Identical, but not the same: Intra-site and inter-site reproducibility of fractional anisotropy measures on two 3.0 T scanners
The results are encouraging for multi-centre DTI studies in larger populations, but also illustrate the importance of the image processing pipeline for reproducibility.
Harmonization of multi-site diffusion tensor imaging data
It is shown that the DTI measurements are highly site‐specific, highlighting the need of correcting for site effects before performing downstream statistical analyses, and that ComBat, a popular batch‐effect correction tool used in genomics, performs best at modeling and removing the unwanted inter‐site variability in FA and MD maps.
Looking for the optimal DTI acquisition scheme given a maximum scan time: are more b-values a waste of time?
This study used simulated data to investigate how valuable the use of multiple b-values is, in terms of improving the accuracy and reproducibility of Diffusion Tensor Imaging (DTI) results, and shows that the systematic bias of the estimated scalar diffusion parameters can be minimized by increasing the number of b- values.
Reproducibility and biases in high field brain diffusion MRI: An evaluation of acquisition and analysis variables.
Results show that reproducibility of FA and MD is comparable to that found at lower fields, not significantly dependent on pre-processing and acquisition protocol manipulations, but that the specific choice of acquisition parameters can significantly bias the group measures of FA, MD, axial and radial diffusivities.
How do spatial and angular resolution affect brain connectivity maps from diffusion MRI?
  • L. Zhan, D. Franc, +9 authors P. Thompson
  • Physics, Computer Science
    2012 9th IEEE International Symposium on Biomedical Imaging (ISBI)
  • 2012
Spatial and angular resolution affected the computed connectivity for narrower tracts, but also for the corticospinal tract, and the apparent role of some key structures in cortical anatomic networks was affected.
Reliability in multi-site structural MRI studies: Effects of gradient non-linearity correction on phantom and human data
The image intensity reproducecibility of the human brain data was significantly improved with distortion correction, suggesting that this method may offer improved reproducibility in morphometry studies.
Cross-scanner and cross-protocol diffusion MRI data harmonisation: A benchmark database and evaluation of algorithms
The results show that cross-scanner harmonisation of single-shell diffusion data sets can reduce the variability between scanners, and highlight the promises and shortcomings of today's data harmonisation techniques.