Center for Modeling of Coupled Subsurface Dynamics
Research project

Network Inpainting via Optimal Transport (NIOT)

The NIOT project goal is to develop new mathematical and numerical tools for the reconstruction of networks from corrupted images. Its ultimate goal is to apply these tools in the reconstruction of blood vessel networks from MRI scans.

Picture of brain with blood vessels
The main NIOT goal: reconstructing corrupted blood vessel images using branched optimal transport
Hodneland et al. 2020 Plos Computational Biology: https://doi.org/10.1371/journal.pcbi.1007073

Main content

Funding:EU, Marie Skłodowska-Curie Actions

The precise digital reconstruction of natural networks such as blood vessels or plant roots is crucial to ensure the quality of simulation-driven predictions. However, these structures can often be accessed only via noninvasive imaging techniques, leading to artifacts that compromise the reliability of the data and the derived simulations.

The NIOT project aims to address the problem of reconstructing networks from corrupted images with two goals in mind:

  • defining a robust mathematical formulation of the problem.
  • proving efficient technological solutions

The core idea of the project is to incorporate the most recent advances of Branched Transport theory within a variational image processing method. The reconstructed network is obtained as the density minimizing the sum of the discrepancy with the observed data and a branch-inducing functional.

Our plan is to test our ideas and software in incrementally difficult problems. Our ultimate goal is to reconstruct the corrupted vascular networks in MRI scans of human patients.