EATx HYBRID ‘Advances in molecular imaging’

In this EATx event features HYBRID, an European Commission funded industrial academic initial training network. The ultimate goal of HYBRID is to establish the field of non-invasive disease characterisation beyond the narrower vision of radiomics. Specifically, they seek to adopt a more inclusive approach to image analysis, that considers anatomical, functional and hybrid imaging, together with related clinical non-imaging parameters that are extracted from patients during their work-up.

Registration is open here. Before registering we kindly ask you to read our Privacy Policy.

The session will feature 4 speakers:

Bert Windhorst will provide an introduction on how EATRIS facilitates multi-site collaborations, showcasing a few examples from industry collaboration and infrastructure development.

Claude Comtat will talk about fully quantitative PET/MR imaging in oncology, based on the work of Florent Besson for lung tumours. Hybrid PET/MR systems will reveal their full clinical potential only if we revisit the acquisition paradigm toward fully quantitative multi-parametric imaging and do not replicate what is currently performed on separate PET/CT and MR systems. Admittedly, fully quantitative MR imaging requires more time than in usual clinical practice, but it also offers the opportunity to perform in parallel dynamic PET imaging for parametric PET imaging. This hybridity result in rich and dense information, requiring advanced processing tools to select relevant parameters for a given clinical question; machine learning techniques are ideally suited to address this challenge. The presentation will be illustrated with PET/MR data for lung tumour imaging.

Georg Schramm will talk about recent advances in the application of AI in molecular image reconstruction. Due to the outstanding performance of AI in the field of image processing (e.g. classification, segmentation), application of AI also emerged in the field of image reconstruction and restoration. This talk is a brief overview of recent research focusing on the use of AI in the process of image reconstruction in molecular imaging.

Guilherme Domingues Kolinger will talk about:

• Spatial registration between brain PET and MRI when they are not acquired simultaneously (i.e. images not coming from a PET/MR scanner)
• Voxel-wise brain image comparison: spatial normalisation of images into a common space
• Use of pre-defined regions of interest (from an atlas) into the original brain PET image