Migration versus apparent migration: importance of errors due to positioning variation in plain radiographic follow-up of aortic stent-grafts.

Abstract

PURPOSE To demonstrate the influence of radiographic positioning on the assessment of stent-graft migration using plain radiographs following endovascular abdominal aortic aneurysm repair. METHODS Equations were derived to correct for artifactual stent-graft migration introduced by geometric distortion due to variations in positioning between radiographs acquired at different times. A phantom system was used to validate the equations. RESULTS Errors in stent position increase with (1) the distance of the aortic stent-graft from the midline and (2) differences in radiographic centering points in the craniocaudal direction; other variables have little effect. For typical stent positions, errors are small if the centering changes by <8 cm. Consistent radiographic positioning to within 4 cm on successive imaging studies limits errors to 1.5 mm. Even if artifactual migration is large, the true migration can be reliably calculated to within 2 mm. CONCLUSIONS Artifactual migration due to variation in radiographic centering is not usually clinically significant if care is taken to center radiographs consistently. Radiographs in which artifactual migration may be important are readily identified, and mathematical correction is straightforward.