Tuned aperture computed tomography (TACT) for cross-sectional implant site assessment in the posterior mandible.

PURPOSE

This study was designed to optimize a new radiographic modality known as tuned aperture computed tomography (TACT) for cross-sectional imaging of implant sites in human dry mandibles.

MATERIALS AND METHODS

Five imaging modalities were compared for image quality and measurement accuracy: (1) conventional hypocycloidal tomograms, (2) TACT images reconstructed using the average method produced using a linear x-ray source movement, (3) TACT images reconstructed using the average method produced using a multidirectional x-ray source movement, (4) minimally reconstructed TACT images without a fiducial marker at the site of interest, and (5) minimally reconstructed TACT images with a fiducial marker at the site of interest.

RESULTS

The extended Mantel-Haenszel mean score statistic was used to investigate the influence of modality on subjective image quality. A statistically significant difference for certain types of TACT images and multidirectional tomography (P < 0.0001) was observed. Linear TACT and multmin TACT were rated as significantly better than other image modalities (P < 0.0009), whereas multidirectional tomography was rated as being significantly worse than other radiographic modalities (P < 0.0001). For the quantitative assessment, data were normalized and analyzed statistically through a paired-comparisons t test. For each modality, the accuracy for maximum height and height was significantly different from ground truth (P < 0.05).

CONCLUSION

The qualitative data suggest that visibility of structures important to the choice of implant location and dimension were seen better with certain TACT methods. Quantitative differences from ground truth (actual measurements of the bone-absolute truth) were clinically negligible. TACT appears to offer the potential of superior image quality over the status quo.