Efficacy and Objective Evaluation of Non-echo-planar Diffusion-weighted MRI by the PROPELLER Technique in the Diagnosis of Cholesteatoma

Abstract

　In the diagnosis of cholesteatoma, non-echo-planar diffusion-weighted MRI using the PROPELLER (periodically rotated overlapping parallel lines with enhanced reconstruction) technique has been reported to be useful. However, interpretation of positive findings is subjective and the decision criteria remain unclear. It is necessary to establish an objective standard for evaluation of the signal intensity.
　The aim of this study was to establish whether diffusion MRI using the PROPELLER technique is effective for the diagnosis of cholesteatoma. Furthermore, we investigated whether objective evaluation is possible by comparing the signal intensities and determining which region of brain is the most suitable for such comparison.
　We evaluated 35 ears with a postoperative diagnosis of cholesteatoma for which preoperative evaluation had been performed by MRI using the PROPRLLER technique. We determined a lesion as positive if it showed a higher intensity than the cerebellar white matter. Furthermore, for objective evaluation, we set regions of interest and compared the highest signal intensity of the target lesion with that of the following parts of the brain; cerebellum, cerebellar white matter, temporal lobe, temporal lobe white matter.
　The sensitivity/positive-predictive value were 85.7%, and the specificity/negative-predictive value was 42.9%. The smallest lesion observed was 3.8 mm in diameter. In addition, in 4 of the 35 ears, no lesion could be detected. Objective evaluation by comparison of the highest signal intensities of the regions of interest matched with the diagnosis after surgery in 26 of the 31 ears. The findings suggested that the cerebellar white matter was the most suitable among the 4 regions evaluated for such comparison.
　This method is thought to be useful, especially in cases where macroscopic evaluation of the lesions is difficult.