Detectability and anatomical correlation of middle ear cholesteatoma using fused thin slice non-echo planar imaging diffusion-weighted image and magnetic resonance cisternography (FTS-nEPID).

ABSTRACT

Cholesteatomas show high intensity in diffusion-weighted imaging (DWI). We performed fused thin slice non-echo planar imaging (EPI) DWI and magnetic resonance cisternography (FTS-nEPID) for cholesteatoma patients to increase the detectability of FTS-nEPID for cholesteatoma. The subjects are 77 consecutive patients who underwent FTS-nEPID as a preoperative study (mean age 53.3±21.8, 47 men and 30 women). Otorhinolaryngologists performed the operations. We anatomically classified the middle ear into four portions. A radiologist evaluated the images for cholesteatoma and assessed the anatomical invasive range in four portions using only FTS-nEPID. We classified large cholesteatomas that invaded more than three portions and small ones that invaded less than two portions based on the results obtained from surgery, and calculated the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). For all cholesteatomas with an existing diagnosis, the sensitivity, specificity, PPV, and NPV were 71%, 70%, 94%, and 27%, respectively. In anatomical evaluation, the sensitivity, specificity, PPV, and NPV were 49%, 85%, 77%, and 64%, respectively. For large cholesteatomas with an existing diagnosis, the sensitivity was 86%. In anatomical evaluation, the sensitivity, specificity, PPV, and NPV were 51%, 57%, 88%, and 18%, respectively. For small cholesteatomas with an existing diagnosis, the sensitivity, specificity, PPV, and NPV were 59%, 78%, 92%, and 30%, respectively. In anatomical evaluation, the sensitivity, specificity, PPV, and NPV were 40%, 85%, 60%, and 71%, respectively. FTS-nEPID may be useful for diagnosing cholesteatomas. Further research is needed for anatomical evaluation because there were many false-negative results.