Diagnostic Value of Electrocochleography and Gadolinium-Enhanced Inner Ear MRI for Patients With Meniere Disease

Objective: To compare the diagnostic performance of electrocochleography (EcochG) and gadolinium-enhanced inner ear MRI in patients with definite and probable Menière disease (MD) and to evaluate their clinical applicability. Study design: Retrospective cohort study. Setting: Tertiary referral center. Patients: A total of 694 patients with MD (765 ears) were included. Ears were categorized into the definite group (704 ears), the probable group (61 ears), and the non-MD group (asymptomatic contralateral ears, 574 ears). In addition, 100 normal ears served as the control group. Interventions: Extratympanic EcochG recordings and gadolinium-enhanced inner ear MRI. Main outcome measures: Diagnostic accuracy of EcochG parameters (amplitude ratio, area1 ratio, area2 ratio) and gadolinium-enhanced MRI, evaluated using receiver operating characteristic analysis. Results: MRI demonstrated the highest diagnostic accuracy in distinguishing the definite group from the non-MD group (sensitivity 79%, specificity 99%). In contrast, EcochG parameters, particularly the area2 ratio, showed superior performance in differentiating the probable group from the non-MD group, with significantly higher area under the curve values than MRI ( P <0.05). A stepwise gradient was observed across groups for both MRI and EcochG parameters. However, area ratios did not significantly differ between the definite group and the probable group. There were weak correlations between MRI findings and EcochG parameters. Conclusions: MRI exhibits high diagnostic value and reflects structural disease burden in definite MD, but its sensitivity in early-stage or probable MD remains limited. Area ratio–based EcochG, especially the area2 ratio, appears more sensitive to early functional alterations associated with endolymphatic hydrops. EcochG might serve as a valuable complementary tool for early detection, whereas MRI remains essential for assessing structural involvement.