The Prototype

Overall architecture

Gamma MRI is a combination of an MRI system with additional gamma detection. As a result, the overall architecture is composed of these two main subsystems: low magnetic field MRI and gamma detectors. Two consoles linked by a synchronization mechanism ensure tight integration of both system components. The prototype is completed with a powerful image reconstruction software.

The Magnet

The magnet was custom-built based on the exact specifications provided by the consortium. However, the shims and gradients were built internally by us because we could not find a supplier able to do the work.

Performing MRI / NMR experiments requires excellent field stability which can be achieved only with proper insulation and active monitoring and control of the magnet temperature. For that purpose, we have attached to the magnet heating pads which can warm the whole magnet’s structure.

Mechanical covers for the magnet have been designed and manufactured to serve several purposes:

  • Thermal insulation to protect the magnet from the room temperature variation.
  • Providing support for the longitudinal detectors.
  • Offering a flat support surface that allows for flexible positioning of transversal gamma detectors.

A standard electronics rack houses the magnet electronics, and the MRI server, as well as the power supply and the amplifiers.

The Gamma Source

As part of the GAMMA source, HESSO developed a complete software suite to facilitate the supervision and control of the different components of the system; including:

  • Automatic laser control,
  • Real time control of the temperature of the Laser
  • Control of the Cooling system of the oven for temperature stability.
  • Safety interlocks.

The Detectors

The gamma detectors fit inside openings made inside the magnet structure. To determine the optimal detector features, different simulations have been carried out, with different variable parameters, so as to reach a final optimal design that meets the original requirements in the most effective way.

Two electronic boards have been designed and custom made with the aim of signal readout and processing. The readout boards contain the power supply and signal processing circuit. The prototype design allows the installation of several detectors in transversal and longitudinal positions. The detectors are heat calibrated using proprietary software.

The Data Acquisition Electronics

Hardware components:

The MRI device is organised, in terms of pulse generation hardware, around two main boards: FPGA core board to numerically generate the different pulses at hight speed rate and input/output MRI board to convert the generated waveform in analogical signal.

Software components:

The main user interface of the system is made in Java and runs under Windows to facilitate ease of use and portability. The pulse sequence kernel is designed as a smart graphical editor with an additional Java code layer that allows advanced calculations. It is compatible with common IDEs and integrates git versioning of the sequence.

The Image Reconstruction Software

The image reconstruction software is one of the key elements in the GAMMA-MRI prototype. The reconstruction process aims at determining the number of events originating in voxel j based on the gamma-ray emission time and direction.

Utilizing a maximum likelihood expectation maximization (MLEM) algorithm, it calculates voxel activity, considering sensitivity through a detailed detector simulation. The iterative reconstruction, monitored for convergence, loops over iterations to find the optimum number of events per voxel. The code generates images from simulated data and provides valuable feedback on physical parameters and their impact on image quality. For a consistent simulation a proper description of the gamma-interactions is also included with effects of shielding, background and scatter. In this way, the efficiency for gamma-ray detection is derived, which enters as the sensitivity parameter in the simulations.

A detailed explanation of the image reconstruction process can be found at the link below:

Image Reconstruction Presentation

The Image Reconstruction process uses iterative methods.