Diagnosis of Enlarged Vestibular Aqueduct Using Wideband Tympanometry

Abstract

Background: Wideband tympanometry (WBT) has the potential to distinguish various mechanical middle ear and inner ear pathologies noninvasively. This study investigated the diagnostic value of WBT in the diagnosis of enlarged vestibular aqueduct (EVA). Methods: The absorbance and resonance frequency (RF) of patients with EVA (40 ears, 25 patients) and matched population controls (39 ears, 28 subjects) were compared, alongside receiver operating characteristic (ROC) analysis. Correlations between VA width and RF were also examined. Results: Patients with EVA had higher absorbance at low frequencies (226-917 Hz) and lower absorbance at high frequencies (2520-4896 Hz) compared to controls. The RF of the EVA group was significantly lower versus controls (751 [391-1165] vs. 933 [628-1346] Hz). The ROC analysis revealed area under the curve values of 0.771 and 0.801, respectively, for absorbance and RF. RF had a sensitivity, specificity, positive predictive value, and negative predictive value of 74.4%, 82.5%, 76.7%, and 80.6%, respectively, for diagnosing EVA. In the EVA group, the VA midpoint width (r = -0.334) and VA petrous width (r = -0.402) both significantly correlated with RF. Conclusions: Our findings support the utility of WBT for diagnosing EVA, with RF as the optimal index used.

Keywords: ROC analysis; absorbance; diagnosis; enlarged vestibular aqueduct; resonance frequency; temporal bone computed tomography; wideband tympanometry.

Figure 1

Measuring the VA midpoint and porous width in axial computed tomography images. The VA midpoint width (white line) was measured at the half the distance from VA fundus to its external pore (A). The VA porous width (white line) was measured from the opercular margins to the spots on the posterior temporal bone walls (black line) whose surface was perpendicular to the measurement lines (B).

Figure 2

Mean absorbance curves at peak pressure against frequency group. The error bars represent ±1 standard deviation from the mean.

Figure 3

Comparison of RF between the control group and EVA group. The mean RF was significantly lower in the EVA group than in the control group.

Figure 4

Receiver operating characteristic (ROC) curve analysis for absorbance at 3776 Hz (A). Area under the curve, 0.771 (95% confidence interval, 0.670–0.871). ROC analysis for RF in control and EVA groups (B). AUC, 0.801 (95% confidence interval, 0.712–0.907).

Figure 5

Receiver operating characteristic (ROC) curve analysis for RF in the (A) child group (area under the curve [AUC], 0.800 [95% confidence interval, 0.684–0.979]) and (B) adult group (AUC, 0.831 [95% confidence interval, 0.684–0.979]).

Figure 6

Correlations of resonance frequency with (A) the VA midpoint width (r = −0.334); and (B) the VA porous width (r = −0.402). The line represents the best-fit regression line.