High-resolution flat panel computed tomography (FPCT) is useful for the evaluation of temporal bone pathologies. While radiation exposure remains a concern, efforts have been devoted to reduce dose, while maintaining image quality. We hypothesize that removing the antiscatter grid (ASG) results in decreased radiation exposure, while maintaining diagnostic image quality for the evaluation of superior semicircular canal dehiscence (SSCD).
Ten patients with clinical suspicion for SSCD participated in this prospective study. Two sequential collimated 20-second FPCT acquisitions were performed (first: grid in; second: grid removed) in all patients. Secondary reconstructions were created by manually generating the volume of interest to include the middle ear using a voxel size of 0.1 mm and 512 × 512 matrix. Radiation dose parameters (air kerma (Ka,r) in mGy and dose area product (DAP) in μGym2) were recorded. Three reviewers analyzed images for the ability to diagnose SSCD, to identify the stapes crurae, and to determine if an ASG was present.
The average Ka,r and DAP for the grid-in acquisitions were 246.7 mGy (SD 47.9) and 2838.0 μGym2 (SD 862.8), versus 160.2 mGy (SD 33.2) and 2026.3 μGym2 (SD 644.8) for the grid-out acquisitions, respectively (p<0.001 for both Ka,r and DAP). Radiation exposure was reduced by approximately 30% solely by removing the ASG. All reviewers correctly identified all patients with SSCD (confirmed at surgery), with mean AUC of 0.99 (κ = 0.90).
Removing the antiscatter grid during FPCT imaging of the temporal bones is a simple and effective way to reduce radiation exposure while maintaining diagnostic image quality for the evaluation of SSCD.
*Division of Interventional Neuroradiology
†Division of Neurootology, The Johns Hopkins University School of Medicine, Baltimore, Maryland
Address correspondence and reprint requests to Monica S. Pearl, M.D., The Johns Hopkins Hospital, Bloomberg Building Room 7218, 1800 Orleans Street, Baltimore, MD 21287; E-mail: firstname.lastname@example.org
Data was presented at the 55th Annual Meeting of the American Society of Neuroradiology (ASNR), April 24–27, 2017 in Long Beach, California.
The authors disclose no conflicts of interest.