Detectability of depth information by use of magnification stereoscopic technique in cerebral angiography.

Two types of stereoscopic phantoms were constructed to determine the sensitivity of observers in detecting depth information obtained by means of the magnification stereoscopic technique. One of these, a "plus" phantom, contained perpendicular aluminum wires at the top and bottom of steps of various thicknesses; the observer was asked to identify whether one of the wires was on the top, the bottom, or uncertain in location. The fraction of correct responses versus the height of the step was plotted, indicating the detectability of depth. Fourteen of 20 unselected observers (70%) responded consistently and were considered "good" observers. In general, the greater the stereo shift, the greater the fraction of correct responses. A second phantom, a mesh phantom, contained standard and randomly positioned test wires. The observer was asked to identify the depth position of a test wire with respect to the standard wire if they were in the same plane. It was found that observers can discriminate depth differences of one to two millimeters in magnification stereoscopy, and similar performance for magnification stereoscopic cerebral angiograms is expected.