J P Guenette1, L Hsu2, B Czajkowski2, D B Nunez2. 1. From the Division of Neuroradiology, Brigham & Women's Hospital, Boston, Massachusetts. jpguenette@bwh.harvard.edu. 2. From the Division of Neuroradiology, Brigham & Women's Hospital, Boston, Massachusetts.
Abstract
BACKGROUND AND PURPOSE: Variable head positioning in the CT gantry results in variable and inconsistent temporal bone imaging planes. Our aim was to evaluate whether an automated postprocessing algorithm or an educational intervention with postprocessing by a technologist could result in consistent temporal bone image reformations into planes referenced to the lateral semicircular canal. MATERIALS AND METHODS: Instructions to reformat small-FOV images in planes referenced to the lateral semicircular canal were posted at 12 CT scanner consoles and e-mailed to 65 CT technologists at a single multisite institution. Automated reformatted images were also produced. The angles between the technologist- and automated-reformatted axial image planes and lateral semicircular canal planes were measured. Group differences were calculated with Mann-Whitney-Wilcoxon tests. Differences in homogeneity of variances were calculated with Fligner-Killeen tests. RESULTS: Two hundred ten temporal bones were imaged in 4 months following the intervention. Reformats by technologists were accurate in 87% of the axial and 81% of the coronal planes, with a trend toward improvement with time. Eighty percent of incorrectly reformatted images occurred at off-site, inpatient, and emergency department scanners. The error angle was significantly lower for technologist-reformatted images (median, 4.9°) than for acquisition plane images (median, 14.6°; P = 3 × 10-14) or automated-reformatted images (median, 13.8°; P = 9 × 10-13). The angle error variance was significantly more homogeneous for technologist-reformatted images (P = 3 × 10-8) and automated-reformatted images (P = 1 × 10-5) than for acquisition plane images. CONCLUSIONS: Both technologist and automated reformatting of temporal bone images resulted in significantly less imaging plane variance compared with images reformatted in the acquisition plane, but reformatting by technologists remains necessary at our institution given our preference for standardized planes referencing the lateral semicircular canals.
BACKGROUND AND PURPOSE: Variable head positioning in the CT gantry results in variable and inconsistent temporal bone imaging planes. Our aim was to evaluate whether an automated postprocessing algorithm or an educational intervention with postprocessing by a technologist could result in consistent temporal bone image reformations into planes referenced to the lateral semicircular canal. MATERIALS AND METHODS: Instructions to reformat small-FOV images in planes referenced to the lateral semicircular canal were posted at 12 CT scanner consoles and e-mailed to 65 CT technologists at a single multisite institution. Automated reformatted images were also produced. The angles between the technologist- and automated-reformatted axial image planes and lateral semicircular canal planes were measured. Group differences were calculated with Mann-Whitney-Wilcoxon tests. Differences in homogeneity of variances were calculated with Fligner-Killeen tests. RESULTS: Two hundred ten temporal bones were imaged in 4 months following the intervention. Reformats by technologists were accurate in 87% of the axial and 81% of the coronal planes, with a trend toward improvement with time. Eighty percent of incorrectly reformatted images occurred at off-site, inpatient, and emergency department scanners. The error angle was significantly lower for technologist-reformatted images (median, 4.9°) than for acquisition plane images (median, 14.6°; P = 3 × 10-14) or automated-reformatted images (median, 13.8°; P = 9 × 10-13). The angle error variance was significantly more homogeneous for technologist-reformatted images (P = 3 × 10-8) and automated-reformatted images (P = 1 × 10-5) than for acquisition plane images. CONCLUSIONS: Both technologist and automated reformatting of temporal bone images resulted in significantly less imaging plane variance compared with images reformatted in the acquisition plane, but reformatting by technologists remains necessary at our institution given our preference for standardized planes referencing the lateral semicircular canals.
Authors: Greg Ogrinc; Louise Davies; Daisy Goodman; Paul Batalden; Frank Davidoff; David Stevens Journal: BMJ Qual Saf Date: 2015-09-14 Impact factor: 7.035