Physical principles and technology of clinical PET imaging.

Molecular imaging with positron-emitting radionuclides is playing an increasingly important role in the diagnosis and staging of malignant disease and in monitoring response to therapy. To meet this challenge, significant improvements in the performance of the imaging technology have been achieved in recent years. Such developments are subject to the constraints imposed by the physics of positron emission tomography (PET) and the main objectives in designing or improving PET scanners are to achieve high spatial resolution and sensitivity while maximising the true coincidence count rate relative to contributions from noise processes. Noise contributions in PET include not only statistical effects associated with photon counting but also background processes such as scatter and random coincidences. The recent developments of new, faster scintillators and electronics for PET detectors, as well as statistically-based algorithms that reconstruct fully three-dimensional (3D) PET images in minutes, have dramatically reduced clinical imaging times while improving image quality. A recent advance, the combination of functional imaging and computed tomography (CT) in the PET/CT scanner has further reduced the study duration by eliminating the lengthy PET transmission scan and providing accurate anatomical localisation of functional abnormalities. PET imaging technology has now improved to where a combined anatomical and functional clinical study can be completed in less than 10 minutes--although taking advantage of such high throughput potential will challenge patient management in diagnostic imaging departments. This paper reviews the physical principles underlying PET and summarises the recent developments in PET scanner technology, from the introduction of new PET detectors to the development of the combined PET/CT scanner.