SAFIR Project Timeline

  1. Beginning of 2014

    Initialization of the project.

  2. 2014

    Detector element characterizations and performance estimations

    Monte Carlo simulations for different detector configurations

    Development of baseline design

    First high rate tests

  3. March/April 2015

    Determine the scanner geometry

    Choose photo-sensor type

    Choose crystal type

    Choose readout ASIC

    Determine number of rings

  4. End of 2018

    SAFIR dual ring detector constructed

    First tests in the Bruker 70/30 MR

    Data acquisition and processing software developed

    Reconstruction software

  5. 2019

    Test the dual ring system and analysis of preliminary results

    Construction of SAFIR full detector (10 rings of 8x8 SiPM arrays)

  6. 2020

    Full SAFIR detector delivered

    Testing of final system and analysis of the results

SAFIR Project introduction


The project was conceived in 2013 as a collaboration between scientists from the ETH Institute for Particle Physics (IPP) responsible for the development of the PET insert, the ETH/University of Zurich (UZH) Institute for Biomedical Engineering (IBT) with Prof. M. Rudin responsible for the MRI-related tasks and the UZH Institute of Pharmacology and Toxicology (IPT) with Prof. B. Weber being the main user of the completed PET/MRI system. The PET insert will cover the full range of PET/MRI applications of the IBT facility whilst being custom tailored to the research needs as defined by UZH IPT.

Work on the project comprising a range of tests and simulations, started in the beginning of 2014. In November 2014 the project was named SAFIR (Small Animal Fast Insert for mRi). An intermediate goal, is the completion of the first detector ring by the end of 2015.

Development of a small animal PET insert for MRI

The objective of the SAFIR project is the development of a unique small animal Positron Emission Tomography (PET) insert for a pre-clinical 7T Magnetic Resonance Imaging (MRI) scanner, permitting truly simultaneous PET and MR data acquisition. The development aims at unprecedented temporal resolution allowing dynamic tracer concentration to be visualized within a few seconds, i.e. on time scales similar to those resolved with MRI.

Fast dynamic PET acquisition with short timeframes has not been transferred to pre-clinical small animal systems so far, the very small resolutions needed (in order of a few millimeters) being a challenge for existing scanners. SAFIR not only targets improved spatial resolution but also improved temporal resolution and will thus be unique in the area of pre-clinical PET/MRI instrumentation and open up a new domain of kinetic and correlated tracer concentration modelling.

Key features of SAFIR:

  • Truly simultaneous PET/MR imaging

  • Excellent temporal resolution, order of few seconds

  • Very high activities, up to 500 MBq, ten times more than in preclinical practice

  • Optimized 4D image reconstruction algorithms for quantitative PET parameter extraction

Challenging requirements for SAFIR

  • Fast dynamic acquisition with image acquisition times of the order of a few seconds

  • High sensitivity, comparable with state-of-the-art commercial PET detectors such as the Inveon or ExploreVista systems

  • Spatial resolution of ~2mm FWHM, comparable with commercial systems such as the Inveon or ExploreVista systems

  • True simultaneous acquisition with fast MRI

  • Full compatibility with the Bruker BioSpin 70/30 USR MRI scanner (7T static field)

  • Front end electronics able to cope with the estimated high data rates, with low pile-up and dead-time, and providing a coincidence time resolution (CTR) of less than 300ps FWHM. This will allow for sub-ns coincidence time windows. Existing application specific integrated circuits will be utilized

  • High off-scanner data rates and throughput

  • A data acquisition system able to handle the high data rates originating from the high source strengths of up to 500 MBq, based on off-the-shelf components

  • Fast dynamic 4D image reconstruction software

Collaboration and Partners

The project is conceived as a collaboration between: