RESUMO
OBJECTIVES: To compare the diagnostic processes for the main unilocular radiolucent lesions of the mandible in the presence of the following variables: conventional and digital radiography, specialization of the examiner and type of lesion. METHODS: Twenty-four panoramic radiographs were selected from the archives of the AC Camargo Hospital (São Paulo, Brazil), aiming at comparing the diagnostic processes for similar unilocular radiolucent lesions of the mandible, with the following histopathological diagnosis: six ameloblastomas, six dentigerous cysts, six keratocysts and six traumatic bone cysts. The radiographs were scanned and processed using the Trophy 2000 software. Three specialists, each from four related areas (pathologists, stomatologists, radiologists and oral surgeons), randomly evaluated the radiographs before and after digitalization. RESULTS: The kappa statistic showed a high level of agreement between results obtained using the two radiographic techniques. This means that, in general, the examiners diagnosed the same cases correctly or incorrectly regardless of the method used. CONCLUSIONS: Based on generalized estimating equations, it was concluded that the probability of correct diagnosis does not depend on the kind of lesion, on the radiographic technique or on the specialization of the examiner. In view of the differing opinions of the specialists regarding the diagnostic validity of some software features available and of the results obtained in indirect digital technique, it may be reasonable to reconsider its use for diagnosis of bone pathology.