This is how anisotropic diffusion phantom reference standards add real value - turning subtle, hard-to-spot inconsistencies into clear, actionable information, ensuring that downstream analysis and clinical decisions are based on trustworthy data.

(6/6)

And relying on anatomy alone, such as checking the corpus callosum on an axial slice, is often not enough to confirm that the b-vectors and the image are aligned.

(5/6)

These issues are more common than most people expect, especially when software packages use different orientation conventions (e.g., LAS vs. RAS).

(4/6)

In our white-matter modules, any L-R orientation mismatch between the image and the b-vectors becomes immediately visible, the diffusion signal falls perpendicular to the fiber bundle axes on both diagonal bundles.

(3/6)

PreOperative Performance’s ground-truth phantoms with white matter simulating modules enable a capability that can be missed in routine practice - clear, objective detection of orientation mismatches and diffusion encoding errors.

(2/6)

DM us or email for more information. Details in bio!

#dMRI #neurosky #neurorad #MRI #neuroscience

This demonstrates that the modules behave as a stable reference, useful for multi-site MRI data comparisons, longitudinal QA & validating AI tools that depend on diffusion metrics.

This is one of the building blocks for quality controlled standards in diffusion MRI across sites and vendors.

6/6

Using this analysis, we are able to see that our bundle has some very high FA values. Using the profiles, we can use statistical methods (TOST) to compare the bundles and assess equivalence of the two profiles to a tolerance level of 0.03, a negligible FA difference.

5/6

Whisker plots of FA values in the bundle bins for two versions of the same phantom:
▶️ outliers above,
▶️ box spans 25th-75th percentile, whiskers show range of values,
▶️ yellow lines show median
and,
▶️ profiles of average FA values along the bundle for both phantom versions.

4/6

Here's what you're seeing in the images -

Our phantom’s 90° crossing fiber module with segmented mask overlaid on the large bundle (in AP direction), and a sagittal view of the same bundle and overlay, and…

3/6

It locates the fiber bundles & segments them into various bins along their length. Averaging voxel values in each bin (for diffusion metrics) gives a profile of the metric along the bundle main axis.

The profile can be compared across phantom units or time points, to assess similarity of data.

2/6

Some last scenes from near Marseille’s Vieux Port. Marseille was founded ~ 600BCE by Phoenician sailors and despite modern infrastructure requirements has wonderfully preserved layers of history.

#MRI #diffusionMRI #DKI #HARDI #neuroscience #neurorad #neurosky

(3/3)

As these imaging modalities are sensitive to noise and inter system variability, reference standards such as our anisotropic diffusion phantom can verify repeatability and consistency of measurements, with applications in #biomarker characterisation, and multi-system and longitudinal studies.

(2/3)

#ESMRMB2025 #microstructuralimaging #UHF #MRI #radsky #diffusionMRI