The College of American Pathologists Phantom Series--an assessment of current nuclear imaging capabilities.

Twelve sets of emission phantoms were distributed semiannually to laboratories of nuclear medicine from 1973 through 1978. This survey aimed to create a national peer group within which it might become possible to compare and standardize nuclear imaging technics, and to evaluate and define the current state of the art in a clinical setting. Results were analyzed in terms of sensitivity (percentage of true-positives), nonspecificity (percentage of false-positives), and performance indices (true-positive/false-negative ratios) for conventional gamma cameras and for rectilinear scanners. The chief determinants of target detectability proved to be target-to-surface depth, target diameter, and intrinsic target contrast. Operational variables and instrument type produced relatively minor effect upon subscriber performance. Gamma cameras tended to show slightly more sensitivity than rectilinear scanners, but were somewhat troubled by nonspecificity. Improvements in the intrinsic resolution of second- and third-generation gamma cameras increasingly employed since 1975 showed negligible impact on performance indices. A consideration of the overall survey performance indices. A consideration of the overall survey performance permits the beginning of a quantitative estimate of current detection capability in day-to-day nuclear imaging.