4k16 0001 Mp 11 Tuesday Dec 02 09:25 AM SV-CVIR (v. 21, #1) 0001 (1332) Cardiovasc Intervent Radiol (1998) 21:11 – 16 C ardio V ascular and I nterventional R adiology  Springer-Verlag New York Inc. 1998 Clinical Investigations Clinical Assessment of a New Stereoscopic Digital Angiography System Thierry Moll, 1 Philippe Douek, 1 Ge ´rard Finet, 1 Francis Turjman, 1 Catherine Picard, 2 Didier Revel, 1 Michel Amiel 1 1 CREATIS, associated to CNRS (UMR 5515) and affiliated to INSERM, Service de Radiologie, Ho ˆpital Cardiovasculaire et Pneumologique, BP Lyon Montchat, F-69394 Lyon cedex 03, France 2 General Electric Medical Systems Europe, DEE Department DREAM, BP 34, F-78533 Buc cedex, France Abstract Purpose: To assess the clinical feasibility of an exper- imental modified angiographic system capable of real-time digital stereofluoroscopy and stereography in X-ray angiography, using a twin-focus tube and a ste- reoscopic monitor. Methods: We report the experience obtained in 37 pa- tients with a well-documented examination. The pa- tients were examined for coronary angiography (11 cases), aortography (7 cases), pulmonary angiography (6 cases), inferior vena cava filter placement (2 cases), and cerebral angiography (11 cases). Six radiologists were asked to use stereoscopic features for fluoroscopy and angiography. A questionnaire was designed to rec- ord their subjective evaluation of stereoscopic image quality, ergonomics of the system, and its medical in- terest. Results: Stereofluoroscopy was successfully used in 25 of 37 cases; diplopia and/or ghost images were reported in 6 cases. It was helpful for aortic catheterization in 10 cases and for selective catheterization in 5 cases. In stereoangiography, depth was easily and accurately perceived in 27 of 37 cases; diplopia and/or ghost im- ages were reported in 4 cases. A certain gain in the three-dimensional evaluation of the anatomy and relation between vessels and lesions was noted. As re- gards ergonomic considerations, polarized spectacles were not considered cumbersome. Visual fatigue and additional work were variously reported. Stereoshift tuning before X-ray acquisition was not judged to be a limiting factor. Conclusion: A twin-focus X-ray tube and a polarized shutter for stereoscopic display allowed effective real- Correspondence to: M. Amiel, M.D. time three-dimensional perception of angiographic im- ages. Our clinical study suggests no clear medical interest for diagnostic examinations, but the field of interventional radiology needs to be investigated. Key words: Angiography — Digital imaging — Stereo- fluoroscopy — Technology assessment In natural human vision, perception of depth depends on monocular and binocular factors. In monocular vi- sion, depth cues are mainly provided by prior knowl- edge of a three-dimensional (3D) object, the relative movement of each of its parts, shadowing, and masking of the background area of the scene. In binocular vi- sion, two phenomena lead to depth perception. Firstly, oculomotor tension variation related to convergence of the eyes informs the visual cortex of the location of the target. Secondly, because the two eyes see an object from slightly different vantage points, the two retinal images of that object are slightly different. This bin- ocular disparity provides the essential cue for binocular depth discrimination or stereopsis. Similarly, stereo- scopic X-ray techniques are based upon the acquisition of two X-ray projections of the same object differing by a slight angle so as to create a binocular disparity [1, 2]. This is a well-known technique for the X-ray ac- quisition of a 3D image by exposing an object to two slightly different projections. Stereoangiography may overcome the problem of depth separation and object overlap found on a single projection of a 3D object [1 – 6]. The technology of twin focal X-ray tubes has solved the inherent difficulties at the acquisition step of the radiological process [3]. However, stereofluoroscopic X-ray imaging is still experimental because there are