Fusion of Computed Tomography and PROPELLER Diffusion-Weighted Magnetic Resonance Imaging for the Detection and Localization of Middle Ear Cholesteatoma

JAMA Otolaryngol Head Neck Surg. 2016 Oct 1;142(10):947-953. doi: 10.1001/jamaoto.2016.1663.


Importance: A method to optimize imaging of cholesteatoma by combining the strengths of available modalities will improve diagnostic accuracy and help to target treatment.

Objective: To assess whether fusing Periodically Rotated Overlapping Parallel Lines With Enhanced Reconstruction (PROPELLER) diffusion-weighted magnetic resonance imaging (DW-MRI) with corresponding temporal bone computed tomography (CT) images could increase cholesteatoma diagnostic and localization accuracy across 6 distinct anatomical regions of the temporal bone.

Design, setting, and participants: Case series and preliminary technology evaluation of adults with preoperative temporal bone CT and PROPELLER DW-MRI scans who underwent surgery for clinically suggested cholesteatoma at a tertiary academic hospital. When cholesteatoma was encountered surgically, the precise location was recorded in a diagram of the middle ear and mastoid. For each patient, the 3 image data sets (CT, PROPELLER DW-MRI, and CT-MRI fusion) were reviewed in random order for the presence or absence of cholesteatoma by an investigator blinded to operative findings.

Main outcomes and measures: If cholesteatoma was deemed present on review of each imaging modality, the location of the lesion was mapped presumptively. Image analysis was then compared with surgical findings.

Results: Twelve adults (5 women and 7 men; median [range] age, 45.5 [19-77] years) were included. The use of CT-MRI fusion had greater diagnostic sensitivity (0.88 vs 0.75), positive predictive value (0.88 vs 0.86), and negative predictive value (0.75 vs 0.60) than PROPELLER DW-MRI alone. Image fusion also showed increased overall localization accuracy when stratified across 6 distinct anatomical regions of the temporal bone (localization sensitivity and specificity, 0.76 and 0.98 for CT-MRI fusion vs 0.58 and 0.98 for PROPELLER DW-MRI). For PROPELLER DW-MRI, there were 15 true-positive, 45 true-negative, 1 false-positive, and 11 false-negative results; overall accuracy was 0.83. For CT-MRI fusion, there were 20 true-positive, 45 true-negative, 1 false-positive, and 6 false-negative results; overall accuracy was 0.90.

Conclusions and relevance: The poor anatomical spatial resolution of DW-MRI makes precise localization of cholesteatoma within the middle ear and mastoid a diagnostic challenge. This study suggests that the bony anatomic detail obtained via CT coupled with the excellent sensitivity and specificity of PROPELLER DW-MRI for cholesteatoma can improve both preoperative identification and localization of disease over DW-MRI alone.

MeSH terms

  • Adult
  • Aged
  • Cholesteatoma, Middle Ear / diagnosis*
  • Cholesteatoma, Middle Ear / diagnostic imaging
  • Diffusion Magnetic Resonance Imaging*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Temporal Bone
  • Tomography, X-Ray Computed*