Preclinical Research

The Imaging Research Laboratory is developing a series of preclinical PET imaging systems around the micro crystal element (MiCE) concept. The systems have various design goals from ultra-high spatial resolution (MiCES), low-cost (cMiCE), high detection efficiency (dMiCE) and multimodality (mrMiCE). Our current research focuses on detector designs with depth of interaction positioning capability and compatibility with MR imaging. In addition to the front end detector development, the IRL has developed a high bandwidth, IEEE-1394a based data acquisition system and is currently developing a second generation IEEE-1394b data acquisition system that will support multi-modality PET/MR imaging.

Recent Selected Abstracts:

Count-rate performance of the DSTE PET scanner using partial collimation

Lawrence R. MacDonald, Ruth E. Schmitz, Scott D. Wollenweber, Charles W. Stearns, Alexander Ganin, Robert L. Harrison, Adam M. Alessio, Thomas K. Lewellen, Paul E. Kinahan  
2007 IEEE Nuclear Science Symposium and Medical Imaging Conference
Oct. 29 – Nov. 4, 2007, San Diego, CA

We investigate the use of partial collimation for clinical PET imaging by removing septa from conventional 2D collimators. The goal is to improve the noise-effective count (NEC) rates compared to 2D and 3D scans for clinically relevant activity concentrations.

We evaluated two cases (1) removing 1 of every 2 septa (called 2.5D) and (2) removing 2 of every 3 septa (2.7D).

System performance was first modeled using the SimSET simulation package. The system was modeled in 2D, 2.5D, 2.7D, and 3D. Cylindrical phantoms with different radii were used to investigate imaging performance for different sized patients. The NEMA NU2-2001 count rate cylinder (20 cm dia., 70 cm long) was used, as well as 27 cm and 35 cm diameter cylinders of the same length. 

The 2.5D and 2.7D configurations were the most promising configurations in increasing NEC based on computer simulations. The 2.5D and 2.7D collimators were then built and tested on our clinical PET/CT scanner (Discovery STE, GEHT).

Results of the simulations were compared to measured results for the 2D, 2.7D, and 3D cases. The agreement between measured and simulated data was excellent in predicting count rate trends, with some deviation in absolute value we believe to be due to energy resolution variation, dead-time corrections, and detector response details that are not modeled in the simulation. The measured 2.7D results showed significant increases in NEC compared to both 2D and 3D modes over a clinically relevant range of activities.

Estimating live-time for new PET scanner configurations

Lawrence R. MacDonald, Ruth E. Schmitz, Adam M. Alessio, Robert L. Harrison, Thomas K. Lewellen, Paul E. Kinahan
2007 IEEE Nuclear Science Symposium & Medical Imaging Conference
Oct. 27 – Nov. 3, 2007, Honolulu, HI

We present the derivation of a live-time model for use in predicting count rates from computer simulations of PET scanners. Computer models are frequently used to investigate new PET scanner configurations, but they typically do not incorporate the count-rate degradation caused by electronics processing which is scanner specific. The live-time fraction depends strongly on the photon flux incident on the detector, as well as the energy distribution of the photons. We modeled the live-time of a clinical PET scanner using measured and simulated data and relative single photon fluxes. The model is based on relating activity in the field of view (FOV) to a photon flux (rate and energy) for various configurations, and generalizing the measured live-time to new configurations by associating live-time with an activity in the FOV via relative photon fluxes. Our model used data from a specific scanner, but the approach is generally applicable.

We applied the live-time model to partial collimation on a PET scanner; in particular a scanner with septa positioned between every second or third detector ring (2.5D or 2.7D collimation, respectively). The photon flux was measured and simulated for conventional collimation (2D, 3D), and simulated for partial collimation. These data were used in the model to predict live-time for partial collimation. Variations in measured and simulated photon fluxes lead to an uncertainty in the live-time model that increased with activity in the FOV. The model was validated against measurements in 2.7D. At low activity the model was very accurate at predicting the live-time fraction. The discrepancy between measured and modeled live-time steadily increased to 5% and 25% at 40 kBq/mL in the FOV for singles and coincidence live-time, respectively. The model allowed us to accurately predict the activity concentration at which the 2.7D noise equivalent count rate was maximized for several phantom sizes.