Non-Echo-Planar Versus Multishot Echo-Planar Diffusion-Weighted Imaging in Diagnosis of Suspected Cholesteatoma

Alazzazy, Mohammad Zakaryia; Tantawy, Heba Fathy Ahmed; Yousef, Alshimaa Yousef AbdElghany; Saleh, Mohamed Hesham Saleh

doi:10.21608/zumj.2025.416136.4124

Non-Echo-Planar Versus Multishot Echo-Planar Diffusion-Weighted Imaging in Diagnosis of Suspected Cholesteatoma

Document Type : Original Article

Authors

¹ Professor of Radio-diagnosis department, Faculty of Medicine, Zagazig University

² MBBCH, Faculty of Medicine, Al Azhar University

³ Lecturer of Radio-diagnosis department, Faculty of Medicine, Zagazig University

10.21608/zumj.2025.416136.4124

Abstract

Background:Cholesteatoma is a destructive middle ear lesion that can cause serious complications if not accurately detected, treated. Diffusion-weighted imaging (DWI) is standard noninvasive tool for diagnosis. We aimed to compare diagnostic performance of non-echo-planar (non-EPI) DWI , multishot echo-planar (EPI) DWI in patients with suspected cholesteatoma, using histopathology as gold standard.

Methods:This cross-sectional study included 18 patients with clinically suspected cholesteatoma.All patients underwent both non-EPI,multishot EPI DWI sequences on a 1.5T scanner. Two blinded neuroradiologists independently assessed images, results were compared with intraoperative histopathological findings.

Results:Mean lesion size was 108.1 ± 110.3 mm² with an average maximal diameter of 13.5 mm. Offensive otorrhea was most frequent symptom (61.1%), followed by loss of balance (50%) tinnitus (50%).Non-EPI DWI detected 14/16 histopathology-proven cholesteatomas (sensitivity 87.5%) and falsely classified 2/2 histopathology-negative cases as positive (specificity 0%); overall accuracy 77.8% and PPV 87.5% (p = 0.01). Median lesion size detected was 36.5 mm² (range 6–105). Interobserver agreement was moderate (κ = 0.52).Multishot EPI DWI classified 3/18 cases as positive; relative to histopathology this yielded 3 true positives, 0 false positives, 13 false negatives (2 true negatives): sensitivity = 18.8% (3/16), specificity = 100% (2/2), PPV = 100% (3/3), NPV = 13.3% (2/15), accuracy = 27.8% (5/18) (p = 0.50, not significant). Median lesion size was 45 mm² (range 30–60). Agreement with histopathology was poor (κ = 0.05)

Conclusion:The present pilot study suggests that non-EPI DWI may offer superior sensitivity, diagnostic accuracy compared with multishot EPI DWI diagnosing middle ear cholesteatoma, particularly for small lesions.

Keywords

Main Subjects