Measurements of tooth length in panoramic radiographs. 2: Observer performance.

Observer performance in tooth-length measurements in panoramic radiography has been examined. Sixty-four teeth, evenly distributed between maxillary first molars, second premolars and mandibular first and second premolars, were fixed in plastic moulds. Each cast was radiographed with an Orthopantomograph, twice with steel balls indicating the cusps and apices of the teeth and once without them. One observer measured the radiographic tooth length twice in both radiographs with indicators in order to estimate the true radiographic tooth length and the repositioning error, and twice in the radiographs without indicators. Seven other observers measured the tooth length on the radiographs without indicators, twice applying their own criteria and twice using defined criteria. The accuracy of the tooth-length measurements was calculated by comparing the true radiographic tooth length with the measurements of the seven observers. The intra-observer variation was defined as the difference between two measurements on the same radiograph. The accuracy and the precision of the tooth-length measurements were highly influenced by the observer performance. The mean radiographic tooth length of the seven observers was closer to the true radiographic tooth length when the observers applied the defined criteria. The inter- and intra-observer variation was higher for the measurements of the maxillary teeth (0.3-1.9 mm) compared with the mandibular teeth (0.3-0.9 mm) for most observers. The highest intra-observer variation was found for the palatal root of the maxillary first molar. The intra-observer variation of four observers decreased somewhat when the defined criteria were used; for the other three observers, however, the variation increased.(ABSTRACT TRUNCATED AT 250 WORDS)