The reproducibility of ultrasonic enamel thickness measurements: an in vitro study.

Abstract

The diagnosis of tooth wear is mostly based on visual scoring systems. Recently, methods measuring ultrasonic enamel thickness have been proposed. The aim of this study was to evaluate the reproducibility and evaluate factors affecting variation of ultrasonic enamel thickness measurements in an in vitro set up. Four observers (1-4) in two institutes (G and S) performed ultrasonic enamel thickness measurements at buccal and palatal surfaces of extracted teeth placed in a phantom jaw. A Panametrics 25DL thickness gauge (Panametrics, Waltham MA, USA), with a 15 MHz transducer, covered by a plexi-glass delay line with a tip diameter of 2 mm was used, coupled to a laptop computer. After one week these measurements were repeated. Subsequently four selected teeth were measured 20 times at a fixed buccal site by one observer. Finally, two observers made ink marks with the instrument probe tip on the buccal and palatal surfaces of all teeth. This was repeated after one week. Digital images were made of the marked teeth and analysed to determine variation in positioning. The underlying enamel thickness variation was subsequently estimated. Results were analysed by calculating the 95% range of measurement variation (RMV, 4 x standard deviation) for each experiment. The main reproducibility experiment showed a 95% RMV of 0.79 mm (observer 1G), 0.64 mm (observer 2G), 0.63 mm (observer 3S), 0.71 mm (observer 4S), 0.51 mm (observer 1S). The repeatability experiment showed a mean 95% RMV of 0.26 mm. The probe position variation in the cervical-incisal direction resulted in an estimated thickness 95% RMV of 0.10 mm (1G) and 0.09 mm (2G). It was concluded that, due to measurement variation, thickness changes of less than 0.33 mm cannot be detected reliably. Probe positioning and a poor repeatability were responsible for only a small part of the total variation. Hence other factors influencing variation in enamel thickness are still to be identified.