A DYNAMIC PHANTOM FOR GENERATION OF CONTRAST BOLUS CURVES FOR THE EVALUATION OF DIGITAL SUBTRACTION ANGIOGRAPHY SYSTEMS

Digital subtraction angiography (DSA) systems are now in use in many r adiological institutions for an increasing number of indications. This emphasises the need for evaluation and optimisiation of the technique s used. Phantoms constructed for evaluation of conventional X ray equi pment are generally not suitable for DSA units. A further complication arises because of the fundamentally different methods used to obtain an enhanced visualisation of the contrast-filled vessels. The most com mon method is the mask-mode subtraction, one frame, the mask, is store d in the memory before arrival of the iodine contrast bolus and then s ubtracted from the following frames in the study. Matched filtering is a post-processing method and the result of the filtered image sequenc e is one single frame. For recursive filtering two low pass filters ar e combined to form a band pass filter which will enhance the frequency content of the contrast bolus and suppress the background anatomy, pa tient movements and noise. A prerequisite for the evaluation of system s based on filtering methods is that the phantom mimics the clinical s ituation. Most DSA phantoms are static phantoms designed primarily for systems based on mask-mode subtraction. Dynamic phantoms are usually based on flow of contrast media in an artificial circulatory system. Q uantitative evaluations of, e.g., signal-to-noise ratio, might be a pr oblem with these phantoms because of variations in contrast flow and t he simulation of bolus curves with a different mean transit times requ ire several measurements. A solid dynamic phantom has therefore been c onstructed that allows evaluation of bolus curves with several differe nt transit times from one singal measurement. The phantom consists of a Perspex plate with bolus curve shaped profiles of silicone rubber pl aced in concentric circles at different radii from the centre. When th e phantom is rotated in the image plane during the exposure the profil es which all have the same length will simulate iodine contrast flows with different transit times. With this phantom all the clinically rel event transit times can easily be simulated and DSA units can be evalu ated independently of which method is used for subtraction of the back ground.