A UK survey of nuclear medicine imaging performance using the TransBone anthropomorphic phantom

In March 1993, the British Nuclear Medicine Society was approached and agreed to take part in an international nuclear medicine quality survey organized jointly by the World Health Organization and the International Atomic Energy Agency. The study involved the use of a novel anthropomorphic transmission phantom, which contained a number of abnormalities of varying contrast. The UK part of the survey was undertaken during April and May 1994. A total of 14 departments took part in the study, using a total of 25 gamma cameras. For each gamma camera tested, imaging was performed using the same camera set-up, imaging and hard copy parameters normally used for recording bone images. The test image was then reported by the usual clinical image reporter according to a fixed protocol and a 4-point scoring system. The results varied widely, with 62% of all reportable locations using all four possible scores between the centres being tested. The sensitivity and specificity ranged from 40 to 80% and 59 to 100% respectively. Calculated ROC areas ranged from 65 to 94% with a mean of 80%. The survey methodology means that a relatively large number of factors affect the overall results, of which the main factor appears to be reporting accuracy.