Double images in plain film radiography: a motion artifact.

D. Marinelli,W. Miller
DOI: https://doi.org/10.2214/ajr.147.6.1312
1986-12-01
Abstract:Movement of a patient during X-ray exposure causes blurring of the resulting image. How motion can lead to two or more discrete images is perhaps less apparent. We describe how a radiograph of an oscillating structure can cause an apparent double image. Failure to appreciate the origin of this artifact can lead to the erroneous conclusion that a double exposure has occurred, resulting in unnecessary reexamination. Figure 1 is a detail of a lateral chest radiograph obtained with a phototimed exposure limited to 0.8 sec. The clip on the aorta marks the proximal anastomosis of a bypass graft. The tip of the pacemaker wire is in the right ventricle, and the aortic valve frame is in the root of the aorta. All three metal objects have indistinct margins due to motion. In addition, each object has produced two discrete images in the radiograph. Our hypothesis is that these objects are oscillating because of cardiac motion, that they are momentarily stationary at both extremes of their excursion, and that the exposure is long enough to show them at both extremes. In a patient with a heart rate greater than 70 cycles/mm, the average cycle lasts less than 0.86 sec. An exposure time of this magnitude would not be unusual when patient is large or when portable equipment is used. To test our hypothesis, we constructed a simple pendulum from an angiographic wire, using a hemostat as a weight. We adjusted the length to yield 70 cycles/mm and obtained a radiograph with a 0.8-sec exposure time. No attempt was made to initiate the exposure during a particular phase of the pendulum’s swing. A detail of the radiograph is shown in Figure 2. Two discrete images of the wire and weight can be seen, with a blurred density in between. The motion of the pendulum can be represented by a sine wave in a time-displacement curve, as in Figure 3. When the pendulum passes through the midpoint of its swing, the velocity is maximum. At the extremes of excursion (represented by points A and B in Fig. 3), the velocity is 0. A radiograph of an oscillating object will always give an image of the object at both extremes of excursion if the exposure time is greater than or equal to one cycle time. If the exposure is between 1/2 and i cycle time, the image will always include one of the stationary points and will sometimes include both stationary points, depending on the initiation of exposure. For example, an exposure from t1 to t2 in Figure 3 will include both extremes of excursion, even though t2 minus t1 is less than the time necessary for a complete cycle. It follows from this discussion that a double image is more likely to be formed when exposure times are longer. We have observed the double-image artifact more often on portable and lateral radiographs and on radiographs of large patients. Aadiographs obtained with shorter exposure times are more likely to show a single blurred image. Once the double-image artifact is recognized, it can easily be differentiated from a double exposure caused by equipment malfunction, and repeat examinations can be avoided.
What problem does this paper attempt to address?