PURPOSE: To compare the sensitivity and specificity of detection, and accuracy of localization, of small steel intraocular and episcleral foreign bodies, using conventional axial and helical computed tomographic scanning in an experimental model. METHODS: Small steel foreign bodies ranging in size from 0.048 to 0.179 mm3 were placed in intraocular and episcleral locations in eye bank eyes mounted in the orbits of a human skull and scanned using helical and conventional axial techniques. Helical scanning was performed using 1-mm and 3-mm thick sections. Conventional axial scanning was performed using 3-mm thick sections. Images were reviewed by masked observers to determine sensitivity, specificity, and accuracy of localization for each imaging method. RESULTS: Steel foreign bodies as small as 0.048 mm3 were detectable with each scanning protocol. Although the helical scans appeared to provide higher levels of sensitivity compared to conventional axial scanning, the difference in outcome between the scan types was not statistically significant. Sensitivity was dependent on the size of the foreign body and ranged from 45% to 65% for the smaller ones (< 0.06 mm3) to 100% for the larger ones (> 0.06 mm3). Multiplanar reformatting of images was helpful in achieving optimal accuracy. CONCLUSION: In an experimental model of steel intraocular foreign body, helical computed tomographic scanning provided images of high quality similar to that of conventional axial scanning.
PURPOSE: To compare the sensitivity and specificity of detection, and accuracy of localization, of small steel intraocular and episcleral foreign bodies, using conventional axial and helical computed tomographic scanning in an experimental model. METHODS: Small steel foreign bodies ranging in size from 0.048 to 0.179 mm3 were placed in intraocular and episcleral locations in eye bank eyes mounted in the orbits of a human skull and scanned using helical and conventional axial techniques. Helical scanning was performed using 1-mm and 3-mm thick sections. Conventional axial scanning was performed using 3-mm thick sections. Images were reviewed by masked observers to determine sensitivity, specificity, and accuracy of localization for each imaging method. RESULTS: Steel foreign bodies as small as 0.048 mm3 were detectable with each scanning protocol. Although the helical scans appeared to provide higher levels of sensitivity compared to conventional axial scanning, the difference in outcome between the scan types was not statistically significant. Sensitivity was dependent on the size of the foreign body and ranged from 45% to 65% for the smaller ones (< 0.06 mm3) to 100% for the larger ones (> 0.06 mm3). Multiplanar reformatting of images was helpful in achieving optimal accuracy. CONCLUSION: In an experimental model of steel intraocular foreign body, helical computed tomographic scanning provided images of high quality similar to that of conventional axial scanning.