Tuned-aperture computed tomography for detection of occlusal caries

Objectives: To evaluate the use of tuned-aperture computed tomography (TACT(R)) for the detection of primary occlusal caries. Materials and methods: The material comprised 76 extracted posterior teeth, 51 with caries, Radiographs were recorded both with the Sens-A-Ray (Regam Medical Systems Int. AB, Sundsvall, Sweden) digital radiographic system and with Kodak EktaSpeed Plus film (Eastman-Kodak Co., Rochester NY, USA) employing an Orthopantomograph OP 100 (Instrumentarium Imaging, Tuusula, Finland) and a Prostyle Intra (Planmeca, Helsinki, Finland) dental unit respectively. The Sens-A-Ray radiographs were used to construct TACT(R) slices and TACT(R) iterative slices. The teeth were subsequently sectioned in 300 mum thick slices. Microradiographs were exposed and analysed with respect to the true presence of lesions. Seven observers subjectively evaluated the radiographs and ROC analyses performed. Conventional radiographs were compared with TACT(R) images by means of the area under the ROC curves, A,. Paired t-test was used to compare A, values. Interobserver agreement was evaluated using the Kendall coefficient and the Friedman Anova. Results: TACT(R) radiographs were significantly better than conventional radiographs for diagnosing all types of occlusal caries combined (P-TACT slices = 0.02. P-TACT iterative slices = 0.01). However, neither TACT(R) system was significantly better than film for enamel and dentinal caries separately. Observer agreement was moderate. However, observers demonstrated significant systematic differences in their readings (P < 0.001). Conclusion: TACT(R) may be a feasible method for diagnosing primary occlusal caries.