Report from SNM: Cherry plots path for PET scanner improvements

June 8, 2006

The future of PET instrument design is likely to follow the pattern of past developments, with incremental improvements to sharpen resolution and increase sensitivity to physiological aspects of disease.

The first simultaneous fusion PET/MR image of a mouse was acquired in May 2006 by Jinji Qi, Ph.D., and graduate student Ciprian Canta. Similar work is under way at the University of Tuebingen. (Provided by Dr. Simon Cherry)

The future of PET instrument design is likely to follow the pattern of past developments, with incremental improvements to sharpen resolution and increase sensitivity to physiological aspects of disease.

During the opening session of the 2006 Society of Nuclear Medicine meeting in San Diego, Simon Cherry, Ph.D, identified six key events in PET scanner development:

  • Whole-body imaging. This key concept made possible applications of PET in oncology and opened the door to reimbursement.

  • 3D PET. This permitted a fivefold improvement in sensitivity compared with 2D imaging with same instrument. Collimating sectors were removed from the machines, allowing detectors to communicate with one another, and better reconstruction algorithms were written. 3D PET rapidly redefined terms for neuroimaging.

  • Advanced scintillators. Better scintillating materials, especially lutetium oxyorthosilicate, raised photon count rates and spatial resolution.

  • Iterative reconstruction. Using iterative reconstruction instead of back projection dramatically improved image quality.

  • PET/CT. Hybrids revolutionized PET practice with dramatically improved sensitivities and specificities for lesion detection and the opportunity to use CT for attentional correction.

  • Small-animal imaging. MicroPET has become an essential component of preclinical imaging for molecularly based drug treatment, creating a bridge for translational research by applying these protocols to human clinical practice.

Cherry, director of the Center for Molecular and Genomic Imaging at the University of California, Davis, then focused on three opportunities to improve PET: time-of-flight PET/CT, various strategies pushing the current 2-mm limits of PET spatial resolution to smaller scale, and simultaneous PET/MRI performed on dedicated whole-body human and small-animal scanners.

The physics of positron emission and annihilation create noise responsible for limiting the spatial resolution achieved with PET/CT cameras, Cherry said. Time-of-flight PET concentrates reconstruction round the few centimeters where positron annihilation occurs, reducing noise and improving resolution. Although TOF-PET was first explored in the 1980s, the higher efficiency of lutetium oxyorthosilicate scintillators and improved reconstruction techniques have made its practical application possible.

Smaller detectors are improving small-animal image resolution. The 1.5-mm minimum detector size now employed contributes to the 1-mm maximum resolution possible on current platforms. LSO detectors as small as 0.05 mm have been developed, and designers are working to configure electronics to connect over 100,000 of these individual devices in an array.

PET/MRI is progressing toward reality. Cherry's lab is developing a hybrid small-animal platform based on a Bruker 7T MR scanner and custom PET insert using positron-sensitive avalanche photo diodes. The first in vivo mouse images from the system were produced last month. University of Tübingen researchers have also demonstrated in vivo imaging from their small-animal PET/MRI system. Siemens' plans for marrying a PET insert to a high-field MR head scanner make its application to humans a near certainty.

Innovations come from understanding the limitations of current technologies, Cherry said. The sensitivity of the center of the field-of-view for whole-body PET may be 5%, but the average over the entire FOV is 2.5% and only 0.3% over the entire body.

Fanciful solutions have been proposed to solve this problem. A true whole-body PET with detectors covering the entire body, eliminating the need for a bed unit, would raise the sensitivity (along with scanner cost and complexity). For whole-body PET/MRI, the combination of true whole-body PET with a extremely long axial FOV MR scanner could be a solution.

As with earlier scanner designs, time will dictate whether FOV, detector miniaturization, or PET/MRI will influence the standard for future PET instrumentation, Cherry said.

Animal imaging emerges onto the research agenda

Innovative component allows development of PET/MR hybrid

Design changes and availability spark PET/CT revolution

Engineers explore the limits of PET/CT imaging