Novel Magnetic Resonance Imaging-Based Method for Accurate Diagnosis of Meniere's Disease

Abstract

Background: Pathologically, Meniere's disease symptoms are considered to be associated with endolymphatic hydrops. Examinations revealing endolymphatic hydrops can be useful for accurate Meniere's disease diagnosis. We previously reported a quantitative method for evaluating endolymphatic hydrops, i.e., by measuring the volume of the endolymphatic space using three-dimensional magnetic resonance imaging (MRI) of the inner ear. This study aimed to confirm the usefulness of our methods for diagnosing Meniere's disease. Here, we extracted new explanatory factors for diagnosing Meniere's disease by comparing the volume of the endolymphatic space between healthy volunteers and patients with Meniere's disease. Additionally, we validated our method by comparing its diagnostic accuracy with that of the conventional method.Methods and Findings: This is a prospective diagnostic accuracy study performed at vertigo/dizziness centre of our university hospital, a tertiary hospital. Eighty-six patients with definite unilateral Meniere's disease and 47 healthy volunteers (25 and 33 males, and 22 and 53 females in the control and patient groups, respectively) were enrolled. All participants underwent 3-Tesla MRI 4 h after intravenous injection of gadolinium to reveal the endolymphatic space. The volume of the endolymphatic space was measured and a model for Meniere's disease diagnosis was constructed and compared with models using conventional criteria to confirm the effectiveness of the methods used. The area under the receiver operating characteristic curve of the method proposed in this study was excellent (0.924), and significantly higher than that derived using the conventional criteria (0.877). The four indices, sensitivity, specificity, positive predictive value, and negative predictive value, were given at the threshold; all of these indices achieved higher scores for the 3D model compared to the 2D model. Cross-validation of the models revealed that the improvement was due to the incorporation of the semi-circular canals.Conclusions: Our method showed high diagnostic accuracy for Meniere's disease. Additionally, we revealed the importance of observing the semi-circular canals for Meniere's disease diagnosis. The proposed method can contribute toward providing effective symptomatic relief in Meniere's disease.