Unique Selling Points


EXPERIENCE

  • >15,000 ASL scans (and 3D T1w, 3D FLAIR), from different populations, different MRI scanners/vendors

  • Closely connected to standardization efforts (OSIPI, ASL-BIDS, RSNA QIBA, NEMA DICOM)

  • Continuous development from both clinical/physiological and physics/engineering perspectives

  • Both graphical user interface (GUI), command-line interface, Docker, tested on Linux, macOS, Windows, laptops and high-performance servers; cloud-solution in development with mediri

PROCESSING

  • Complete ASL processing pipeline including geometric distortion correction, motion correction, partial volume correction, M0 processing

  • Quantification options for all common ASL labeling types, including single-PLD, multi-PLD, look locker, Hadamard, multi-echo, also through FSL BASIL

  • Optimized structural processing including CAT12 segmentation & non-linear registration, SPM12 longitudinal registration

  • Population-specific and vendor/sequence-specific templates and atlases

  • Fully automatized batch processing and parallelization

  • Acquires anatomical information (volume/atrophy), vascular information (spatial CoV) in addition to localized tissue perfusion.
    Ability to differentiate between structural, vascular, and localized tissue perfusion effects

  • Output results in any atlas possible, including by default MNI structural, Harvard-Oxford, or Hammers atlas

  • Vascular artifact reduction

QC

  • Internal database of all different ASL flavors from all scanners on which all processing improvements are continuously tested

  • provide an overview of present and processed files, as well as files that were not imported or not processed

  • Automatic QC and unit-testing of each pipeline component, including comparison with reference images

  • Visual QC for each processing step, all in MNI space for easy comparison between studies

  • Estimation and correction of additional between-sequence effects such as effective spatial resolution

  • checks DICOM and BIDS values such as TR, TE, and scale slopes across individuals, automatically picking up on protocol changes, software updates, etc., that can lead to unwanted variance in the statistical analyses

  • automatically checks for left-right flips

  • Matlab and SPM12 code changes for improved reproducibility