Pöschl Reformations Created from High-Resolution Noncontrast Enhanced CT Head Exams Can Be Used to Detect and Classify Superior Semicircular Canal Abnormalities.

Objective: To compare the diagnostic capability of Pöschl reformations created from temporal bone CT (TBCT) and high-resolution noncontrast CT head exams (HR-NECTH) to detect and classify superior semicircular canal (SSC) abnormalities.

Study design: Retrospective case review.

Setting: Tertiary referral center.

Patients: Individuals with normal and abnormal SSC who received TBCT and HR-NECTH.

Interventions: Pöschl reformations of each patient's temporal bones (TB) were created using TBCT and HR-NECTH. Screenshots displaying the SSC were randomized into an interpretation test sent to neuroradiologists who interpreted the SSC as normal/abnormal and classified abnormalities as thinning (SST)/dehiscence (SSD).

Main outcome measure: Analysis of responses to interpretation test, interrater reliability, and sensitivity, specificity, and positive/negative predictive values of Pöschl reformations from TBCT and HR-NECTH.

Results: Fourteen patients were enrolled, 28 TB were reviewed. Sixteen TB demonstrated SST/SDD and 12 were normal as per gold standard TBCT. Interpretation test displaying screenshots of the SSC, in a randomized, blinded fashion, was completed by four neuroradiologists. Analysis of TBCT and HR-NECTH Pöschl reformation interpretations yielded no statistically significant difference in proportion of true/false-positive/negative responses (χ2 = 3.37, p = 0.83), similar and substantial interrater reliability (k = 0.78 vs. 0.73, respectively), and similar sensitivity, specificity, and positive/negative predictive values (0.86, 1, 1, 0.84 vs. 0.78, 0.96, 0.96, 0.77, respectively). Neuroradiologists also correctly classified SST/SSD on TBCT and HR-NECTH Pöschl reformations at a rate of 70.9% vs. 78.0%, respectively.

Conclusion: Pöschl reformations created from HR-NECTH demonstrated similar diagnostic capability as those created from TBCT to detect and classify SSC abnormalities.