RESUMO
OBJECTIVE: To evaluate observer faciolingual depth and vertical depth reading errors when using various X-ray beam array geometries to make basis images for tomosynthetic reconstruction using tuned aperture computed tomography (TACT). MATERIALS AND METHODS: Tissue-equivalent models were constructed to replicate the position of dental implants in relation to simulated mandibular canals. X-ray beam geometries used to acquire the basis images for TACT(TM) integration included horizontal linear, vertical linear, and symmetric and asymmetric conical arrays. Twenty-one dentists trained in the use of TACT acted independently as observers. Tasks included: (1) determination of the relative position of the implant in relation to the simulated canal; and (2) measurement of the vertical depth and faciolingual (lateral) depth dimensions between these two structures. As the study did not involve repeated measures (only one measure was obtained from each observer on each of the two dependent variables), data for faciolingual depth and vertical depth reading errors were analysed using a one-way analysis of variance (ANOVA) with Tukey's honestly significant different (HSD) procedure. RESULTS: Errors in determining the relative position of the implant to the simulated canal were most frequent when the linear horizontal projection geometry was employed for producing basis images (57% error for model #2 where the implant was lingually placed). The mean measurement errors for TACT images constructed using the various different projection arrays depended both on the structural relationship of anatomic features and the employed beam array geometry. CONCLUSION: Conical beam arrays are preferred over linear beam arrays for constructing basis images used with TACT for the purpose of correlating the position of a mandibular dental implant in relation to the mandibular canal. They more consistently allowed the observers to establish a measurement of the faciolingual relationship of the implant to the canal. For vertical depth measurement of the relationship of a mandibular dental implant to the mandibular canal, TACT does not appear to have any advantage over individual simple transmission radiographic images.
