News

 

AN IMAGE ANALYSIS COMPETITION FOR THE BRAIN PET COMMUNITY

For many years PET centres around the world have developed and optimised their own analysis pipelines, including a mixture of in-house or independent software and have implemented different modelling choices for PET image processing and data quantification. As a result, many different methods and tools are available for PET image analysis.

This Grand Challenge aims to provide an opportunity to apply different approaches to the same data in order to assess the performance and consistency of PET modelling approaches on a common data set. This aligns with current interests in the community for sharing of data sets and investigating alternative analytical strategies.

Aim of the challenge

This competition is aimed at evaluating different PET analysis tools on a common data set to explore the performance of analytical tools developed by individual research groups.  The particular challenge is to identify areas of receptor binding changes in a PET radioligand neurotransmission study. The results of the Grand Challenge will be presented at the NRM2018 meeting.

General Principles

  • A simulated data set has been generated for a set of dynamic PET scans where the ground truth is known.
  • Participants will be provided with a number of scans pre- and post-challenge and the goal will be to identify areas and magnitude of change in receptor availability.
  • Participants will need to process the dynamic PET scans with appropriate image processing and kinetic analysis methods.
  • Evaluation criteria will be clearly identified and provided to participants. Individual performance from all entrants will be discussed only anonymously.

How to take part

Registration is not open yet. If you want to receive updates register your interest.

 

To learn more about NRM 2018 please visit their website.