Implementation of the First Triple Modality System Model in STIR

The combination of diagnostic and therapeutic radionuclide is increasingly being used to provide a Theragnostic approach to cancer treatment. In this approach positron emission tomography (PET) and single photon emission computed tomography (SPECT) are used to estimate the absorbed dose to a specific organ and to monitor therapy. Triple modality imaging tools including PET, SPECT and CT have been released in the market. To take advantage of such devices and allow the investigation of their benefits, we extend our implementation of the Mediso AnySan SPECT/CT with the AnyScan PET model. We used the open source software for tomographic image reconstruction (STIR 4.0), and implemented functionalities to read, process, and reconstruct the data from the Mediso AnyScan SCP scanner. In particular, a framework to estimate normalisation, attenuation, randoms and scatter corrections has been created. Two phantom datasets were used. A uniform cylindrical solid Ge-68 phantom was used to validate the software, and a solid Ge-68 NEMA phantom with spherical inserts was used to demonstrate the quality of the image reconstruction. The effect of the aforementioned corrections was studied qualitatively, whereas visual comparison was carried out between the reconstructed images with the vendor software and with STIR. Moreover, we demonstrated the feasibility of PET image reconstruction using the CT and iterative PET image estimates as prior information. Finally, a run time performance study showed that, when using multiple cores, the speed of all the algorithms is within 20% difference from the MLEM, thanks to the recent parallelisation of the kernel estimation in STIR 4.0. Other functionalities, such as decay correction and resolution modelling, are work in progress and a first demonstration of the reconstruction making use of all the triple modality information will be included in the future.