Recently, our team has developed a novel calibration method by which the light response functions (LRFs) of the photosensors in a position sensitive scintillation detector can be iteratively calculated from the standard flood field calibration (the detector is uniformly irradiated by an uncollimated radioactive source). There is no strict requirement on the irradiation uniformity or the energy spectrum, making the calibration process straightforward and quick. We have also confirmed, using both simulated and experimental data, that this method can be successfully applied to a clinical gamma camera of classical design. In the course of the previous FCT project, we have developed an integrated software package specifically targeting scintillation cameras and providing a set of easy-to-use tools for simulation and reconstruction of scintillation events in cameras of configurable geometry. Using simulated data we have shown that for a typical medical gamma camera geometry, the iterative method produces the LRFs that are very close to the photosensor response used to simulate the flood field data. To obtain experimental confirmation, we upgraded a decommissioned clinical gamma for list mode data acquisition with which images of a bar phantom were obtained showing little to no distortion over the whole field of view. We were also able to continuously monitor variations of the PMT gain from acquired data in real time. In this project we plan to upgrade the acquisition system of the prototype of the self-calibrating clinical camera, constructed during the previous project, in order to improve acquisition rate and photon integration time. This will allow to reach characteristics expected from a commercial model and help to attract potential investors. As a parallel line of work we intend to apply a similar calibration technique and implement real-time statistical position reconstruction to compact gamma cameras. For this, we plan to combine our iterative approach with machine learning methods to develop hybrid technique(s) for reliable LRF reconstruction in rectangular crystals and integrate it into our software package. We will develop and optimize design of compact gamma camera in order to maximize the undistorted field of view, achieve highest possible spatial resolution and to attempt accurate depth of interaction reconstruction. Two prototypes are envisaged: a thin crystal camera for use with a parallel hole collimator and a thicker camera with a pinhole collimator for small animal SPECT.

Photos

Contacts

Group Leader:  

Vladimir Solovov