Analysis of the accuracy of high-resolution computed tomography techniques for the measurement of stapes prostheses

Abstract

Severe cochleovestibulopathy is a complication after stapes plasty and CT may be needed to assess intravestibular intrusion depth of the stapes piston. The accuracy of modern CT systems in this task has not been established. Stapes plasty was performed on six cadaver ears, and Soft CliP titanium pistons were inserted. Specimens were then examined using three different CT systems: single source 64-row, dual source 64-row and flat-panel computed tomography (FD-CT). Piston length, width and intravestibular intrusion were measured by four independent observers. Temporal bone dissection was then performed and vestibular piston intrusion measured microscopically. Piston dimensions were overestimated by all observers and CT systems. Intravestibular penetration as assessed by CT was consistently greater than the value found on cadaveric anatomical dissection. On average, the CT measurement of width was 0.176 mm (44%) greater, length 0.125 mm (6.25%) greater and intrusion 0.333 mm (39.2%) greater than the anatomical measurements. FD-CT showed better image quality and partly less bias in several measurements, but penetration depth was still variably overestimated. CT measurement consistently overestimated intravestibular piston dimensions and vestibular intrusion. Modern temporal bone imaging systems are not yet able to depict the stapes piston position with a sufficient degree of accuracy.