A computerized method for measuring relative coronary arterial stenosis by cinevideodensitometric analysis of 35 mm coronary arteriograms was developed and validated. Video images of projected coronary arteriographic frames were digitized into a 512 x 512 matrix (256 gray levels) by computer analysis that compared integrated contrast density measured over stenotic and normal arterial segments after background subtraction. Pixel density was 70 to 80 pixels/mm2 actual area. In phantom studies performed on plexiglass cylinders, cinevideodensitometric measurements correlated linearly with concentration of contrast medium (r = .99), with cross-sectional areas (r = .99) of contrast-filled cylinders 1 to 4 mm in diameter over a wide range of contrast concentrations (25% to 100%), and with relative stenosis of eccentric lesions in the cylinders (r = .99, SEE -3.9%). In another. Second, an estimation of relative stenosis based on minimal cross-sectional area is a more accurate indicator of blood flow than is that based on the degree of relative narrowing expressed in terms of relative diameters,9-U0 which is the traditional method for interpreting coronary arteriograms subjectively.The fact that there are no techniques for accurately grading the severity of coronary atherosclerotic narrowings from coronary arteriograms is a major limitation in the study of ischemic heart disease. DIAGNOSTIC METHODS-CORONARY ARTERY DISEASEAngiograms recorded on conventional 35 mm arteriographic film can be rapidly digitized by computer processing of the projected angiographic image. The projected image is recorded with a vidicon camera, and the resulting video signal is digitized into a 512 x 512 matrix with 255 gray levels, providing quantification of the optical density of each region of the film. The optical density of each arterial segment recorded on the angiographic frame reflects the volume of contrast medium within the arterial lumen, which directly represents the volume of the arterial segment.In this study a method for measuring the cross-sectional area of coronary stenotic lesions by videodensitometric analysis of 35 mm coronary arteriograms was developed and validated in experiments with phantom models and postmortem human hearts. The hypothesis that videodensitornetric analysis of the optical density of contrast medium within an artery reflects the crosssectional area of the arterial lumen was tested. MethodsRadiographic equipment. Coronary cinearteriograms were recorded on 35 mm cine film (Kodak CFR) at 32 frames/sec with an Arriflex camera. The radiographic equipment consisted of a Philips modular generator, an SRN 10/80 x-ray tube, and a trimodal (6, 10, 14 inch) cesium iodide image intensifier mounted on a Poly Diagnost A arm and scanned with a Plumbicon video tube. The focal spot size for coronary cine fluorography was 0.7 mm. The image intensifier was operated in the 6 inch mode, and x-ray exposure time was 5 msec. Due to the large size of the 14 inch image intensifier, the curvature of the input phosphor va...

No abstract

The purpose of the study was to perform phantom studies to assess the impact of computed tomography (CT) system variability on quantitative measurements of contrast enhancement. A phantom containing tubes of contrast material at dilutions of 120, 1:35, 1:50, 1:100 and 1:200 arranged in air or water was imaged using 11 CT systems at 9 institutions. All systems had undergone routine calibration against air and water in accordance with the manufacturers' recommendations. For a given tube voltage, the relationship between the iodine concentration and CT attenuation value on a single system varied by 17 to 24% over 46-48 weeks. The coefficients of variance for iodine calibration factors across different CT systems were 8.9% in air and 5.1% in water. Calibration of individual CT systems for iodine response is required to allow comparison of quantitative measurements of contrast enhancement across different institutions. Using the iodine calibration factor to express contrast enhancement as iodine concentration would facilitate the universal application of diagnostic enhancement thresholds, especially if the necessary calculations were performed by software installed on the CT console.

No abstract

In tomographic medical devices such as SPECT or PET cameras, image reconstruction is an unstable inverse problem, due to the presence of additive noise. A new family of regularization methods for reconstruction, based on a thresholding procedure in wavelet and wavelet packet decompositions, is studied. This approach is based on the fact that the decompositions provide a near-diagonalization of the inverse Radon transform and of the prior information on medical images. Corresponding algorithms have been developed for both 2-D and full 3-D reconstruction. These procedures are fast, non-iterative, flexible, and their performances outperform Filtered Back-Projection and iterative procedures such as OS-EM.