Abstract
To evaluate the accuracy of cone-beam computed tomography (CBCT) images acquired with 40- and 20-s exposures for the detection of surface osseous changes in the temporomandibular joints (TMJ) compared with the gold standard and to examine the interobserver and intraobserver reliabilities of both exposures.
The gold standard was derived from naked-eye inspection of 88 sites in eight TMJs from four dry human skulls. The skulls were scanned by CBCT with 40- and 20-s exposures. Two experienced maxillofacial radiologists interpreted the images according to case definitions of surface osseous changes for the study. The findings were compared with the gold standard findings, and the intraobserver and interobserver reliabilities were examined.
Sensitivities of the exposures were comparable: 53.7, 31.7, and 41.5 % for 40 s versus 31.7, 39, and 34.1 % for 20 s, with overlapping confidence intervals. Specificities were high and comparable: 76.6, 74.5, and 80.9 % for 40 s versus 66, 80.9, and 80.9 % for 20 s, with overlapping confidence intervals. Intraobserver reliability was significantly correlated in 40 (p = 0.0030) and 20-s (p = 0.0001) exposures. Interobserver reliability was significantly correlated in the first 40-s exposure reading (p = 0.0013), but not in the second reading (p = 0.0879), and significantly correlated in the two 20-s exposure readings (p = 0.0016 and p = 0.0097).
Decreasing exposure does not affect detection accuracy of TMJ surface osseous changes. Further studies with larger sample sizes are warranted to assess the effects of exposure time and other factors on CBCT image accuracy.