Digital subtraction angiography applied to peripheral runoff studies

Peripheral runoff studies are one of the most common, yet one of the most painful, of angiographic studies. Digital subtraction followed by image contrast enhancement can help reduce both the pain and the risk associated with angiographic examinations by reducing the amount of iodine contrast media required to view the vasculature being studied. Unfortunately, conventional digital subtraction techniques require one iodine injection for each view. Although pain is reduced, the total amount of the toxic iodine injected in not. In addition multiple injection techniques might increase the total time required to perform the study. Thus, there is great incentive to perform peripheral angiographic studies using digital subtraction via a single injection for multiple views. Single injection, multiple view peripheral angiography possesses many technical difficulties. Patient motion and positioning inaccuracies can significantly reduce the quality of the subtracted images. Although patient motion can be significantly reduced in the periphery by appropriate restraints and instructions to the patients, positioning errors between the mask and contrast frames can lead to uncorrectable misregistration artifacts due to the effects of parallax. Continuously sweeping the imaging chain along the length of the legs might prove advantageous, as it reduces the time wasted while waiting for a stepping table to get the to its next position and stop vibrating. However, a continuous sweep would introduce blurring in the image. To test the feasibility of digital acquisition during continuous sweeping in peripheral runoff studies, a pilot system was built by interfacing a conventional moving table fluoroscopy system to a digital acquisition system and a control computer. Images were acquired at preselected positions while panning the table during the study period. The registration accuracy between the iodine-containing and the mask images was determined from phantom and patients studies. Mis-registrations between the mask and contrast frames of up to 1 to 2 mm were correctable via subpixel linear translations. The experiments with the pilot system also demonstrated that blurring due to the continuous motion is negligible as long as the panning speed and x-ray pulse width were kept within certain bounds.