Learning Embeddings for Atlas Propagation (LEAP)

IXICO's patented LEAP (Learning Embeddings for Atlas Propagation) technology uses AI to provide a customized anatomical atlas for a given MRI image, leveraging IXICO’s large reference database to obtain an optimized segmentation for every given case. LEAP allows us to measure brain volume both cross-sectionally and longitudinally. This provides the following measures: 

  • Global – making measurements on the whole brain
  • Regional – measuring specific local brain regions

LEAP algorithm has been applied to thousands of datasets and provides accurate delineation of key brain structures, while quantifying the volume of up to 137 structures in a single workflow. Using multi-atlas propagation, a flexible and robust method for image segmentation, LEAP is optimized and validated for therapeutic area and region specific segmentations.

In addition to powering our volumetric MRI solutions, LEAP can be deployed in conjunction with other methods such as dMRI, fMRI and PET, enabling the superior accuracy, rigor and efficiency inherent in LEAP algorithms to enhance performances of the paired analyses.

We cannot play this video. The video is embedded from YouTube, your current cookie preferences do not allow the storing of some of the cookies the player will place.

Watch LEAP in action

Want to know more about LEAP and its clinical research applications? Here is our Imaging Science Manager, Richard Joules, explaining how the AI works, its benefits in key therapeutic areas (in this case, Alzheimer's disease), how its accuracy compares to manual analysis, how it can support your clinical trial operations, and more.