VIDEODENSITOMETRIC TECHNIQUES IN DIGITAL SUBTRACTION VIDEOANGIOGRAPHY

Abstract

During the past year, digital subtraction videoangiography has proven to be a powerful tool for intravenous studies as well as enhancing the capabilities of intra-arterial techniques in contrast limited situations. With this imaging technique, it is possible to establish a linear relationship between the subtraction signal and the contrast thickness. The digital nature of the image processing enables convenient retrieval and manipulation of image data. For the above reasons, digital subtraction videoangiography lends itself to videodensitometric measurement of physiological quantities. However, the implementation of videodensitometry capabilities requires extensive care in the processing of imageintensified video signals. Radiation scattering and veiling glare in the image intensified tend to bias the black level for logarithmic amplification and suppress the subtraction signal. In addition, pincushion distortion of the image intensifier may cause a deviation in the digital representation of area. Absolute linearity must also be maintained throughout the image processing, except for the logarithmic amplification. These problems, as well as their solutions, will be discussed in this paper. A technique using real-time digital convolution to remove the scattering component in the image intensified video will be presented. With the necessary corrections, our videodensitometry system has been calibrated and used to estimate canine left ventricular ejection fractions. The possibility of using videodensitometric techniques to characterize the blood flow in vessels and correlate it with the degree of stenosis has also been studied.