Introduction

Interventional radiology is related to the visualization of patient vasculature for the purpose of diagnosing problems such as aneurysms and arterio-venous malformations (AVM). Interventional radiology is also increasingly involved with performing therapeutic procedure (e.g., plugging an AVM), thereby obviating the need for invasive surgical procedures. Since blood has very similar x-ray attenuation properties to those of adjacent soft tissues, vascular structures are not normally seen on conventional radiographs or CT images. However, if a material such as Iodine is added to the flowing blood, the increased x-ray attenuation of the added iodine will permit the blood vessel to be seen. Iodinated contrast medium absorbs (much) more than blood/tissue, and thus would appear whiter on a conventional radiographic image since less radiation would be incident on the film. It is also possible to add contrast agents such as carbon dioxide into the blood stream that will transmit more radiation than the surrounding soft tissues. For images of carbon dioxide contrast obtained with conventional screen/film, the vasculature will look darker than adjacent soft tissues.

The two images in the first column of Figure A (left) illustrate digital images obtained of a patient’s leg following the administration of iodinated contrast. The image was acquired using an x-ray imaging chain that uses an image intensifier to capture the pattern of x-rays that is transmitted through the patient. The image intensifier converts the x-ray pattern into a corresponding light pattern that is captured by a TV camera whose output is digitized. The matrix size of the images shown in Figure A are 1000 x 1000, with each pixel coded using 10 bit that corresponds to 1024 shades of gray. Note that in these raw images, the areas around the patient’s leg look bright because high detected radiation intensities give rise to large digital numbers; normally high digital values would appear bright on a display monitor. Conversely, areas that transmit little radiation will appear dark on raw digital images for the same reason. All digital x-ray imaging modalities behave in this manner, which is opposite to conventional screen/film radiographs where high transmitted intensities have resulted in a darker film image. Accordingly, it is normal practice to digitally invert digital x-rays (i.e., make the dark areas bright and vice versa) before displaying these to radiologists maintain the traditional display paradigm (high radiation levels appear dark and low radiation appear bright).

In the conventional images in Figure A, the blood vessels are difficult to detect because of the overlapping anatomical structure. A mask image may be subtracted from the images that have varying amounts of iodinated contrast material in the blood vessels; the mask image contains the anatomical information, whereas the subsequent images contain both the (same) anatomical data, as well as additional attenuation from the iodine in the vasculature. Subtraction of the mask from subsequent images, as depicted in columns 2 through 4 in Figure A, permits the flow in leg to be visualized as a function of time. Note that removal of the anatomical structures dramatically improves the vasculature visibility. This form of imaging is normally called Digital Subtraction Angiography (DSA), and was first introduced into clinical practice in the late 1980’s. Today, DSA imaging is a mainstay of Interventional Radiology.

Figure A. Clinical example of digital subtraction angiography (femoral) showing the upper region (above) and the lower region (below). In each row, the first image is an example of an normal (i.e., not subtracted) image showing the full patient anatomy as well as the iodinated contrast agent in the vasculature; sequential images to the right show just the vasculature, as obtained by subtracting a “mask” image (no iodinated contrast agent) from sequential images obtained as the iodinated contrast agent progresses down the patient’s leg.

rsna