On-line sliding-window list-mode PET image reconstruction for a surgical pet imaging probe

We present an on-line image reconstruction method for a surgical PET imaging probe. We have been investigating the PET imaging probe in order to detect small tumors or remnants of surgically removed tumors either during surgery or preoperatively. The probe system consists of a segment of a conventional PET scanner and a small, high-resolution PET detector equipped with a position-tracker and collects coincidence events between them. The imaging probe’s proximity to the lesions and the small size crystals in the probe contribute to high spatial resolution. It is anticipated that in the clinical system, the acquired data will be presented to the clinicians or surgeons on line. In order to display the images on line, we use a variant of the list mode OSEM. By consecutively applying a nonoverlapping data window or an overlapping data window on data, we can present intermediately reconstructed images on line. The continuously updated 3D image is re-projected in real-time onto a plane whose orientation is driven by the tracking system. We are interested in implementing the real time PET imaging system using graphic process units (GPUs) and parallel processing. We tested out the algorithm using simulated data.