Computer Assessment of Carotid Intima-media Thickness
Explore the source record for details and available documents.
Engineering topics
Publications and source records attributed to Blankenhorn, D. H..
Explore the source record for details and available documents.
The Cholesterol Lowering Atherosclerosis Study, a randomized angiographic clinical trial, demonstrated the beneficial effect of niacin/colestipol plus diet therapy on coronary atherosclerosis. Outcome was determined by panel-based estimates (viewed in both still and cine modes) of percent stenosis severity and change in native artery and bypass graft lesions. Computer-based quantitative coronary angiography (QCA) was also used to measure lesion and bypass graft stenosis severity and change in individual frames closely matched in orientation, opacification, and cardiac phase. Both methods jointly evaluated 350 nonoccluded lesions. The correlation between QCA and panel estimates of lesion size was 0.70 (p less than 0.0001) and for change in lesion size was 0.28 (p = 0.002). Agreement between the two methods in classifying lesion changes (i.e., regression, unchanged, or progression) occurred for 60% (210 of 350) of the lesions kappa +/- SEM = 0.20 +/- 0.05, p less than 0.001). The panel identified 442 nonoccluded lesions for which QCA stenosis measurements could not be obtained. Lesions not measurable by QCA included those with stenosis greater than 85% that could not be reliably edge tracked, segments with diffuse or ecstatic disease that had no reliable reference diameter, and segments for which matched frames could not be located. Seventy-nine lesions, the majority between 21% and 40% stenosis, were identified and measured by QCA but were not identified by the panel. This comparison study demonstrates the need to consider available angiographic measurement methods in relation to the goals of their use.
A computer method for detection and quantification of atherosclerosis from angiograms has been developed and used to measure lesion change in human clinical trials. The technique involves tracking the vessel edges and measuring individual lesions as well as the overall irregularity of the arterial image. Application of the technique to conventional arterial-injection femoral and coronary angiograms is outlined and an experimental study to extend the technique to analysis of intravenous angiograms of the carotid and cornary arteries is described.
The velocity of sound in excised human and canine arterial tissues is measured in order to serve as a basis for the development and application of ultrasonic techniques for the diagnosis of atherosclerotic lesions. Measurements of sound velocity at different regions of 11 human and six canine aortas were made by a time delay spectrometer technique at frequencies from 2 to 10 MHz, and compared with ultrasonic attenuation parameters and the results of biochemical assays. Sound velocity is found to increase with increasing attenuation at all frequencies, and with increasing collagen content. A strong dependence of sound velocity on cholesterol content or low calcium contents is not observed, although velocities of up to 2000 m/sec are observed in highly organized calcified lesions. A decrease in velocity with decreasing temperature is also noted. It is thus concluded that it is principally the differences in tissue collagen levels that contribute to image formation according to sound velocity.
Computer-enhanced visualization of coronary arteries and lesions within them is discussed, comparing invasive and noninvasive methods. Trial design factors in computer lesions assessment are briefly discussed, and the use of the computer edge-tracking technique in that assessment is described. The results of a small pilot study conducted on serial cineangiograms of men with premature atherosclerosis are presented. A canine study to determine the feasibility of quantifying atherosclerosis from intravenous carotid angiograms is discussed. Comparative error for arterial and venous injection in the canines is determined, and the mode of processing the films to achieve better visualization is described. The application of the computer edge-tracking technique to an ultrasound image of the human carotid artery is also shown and briefly discussed.
A quantitative basis is developed for the medical application of ultrasonic spectroscopy, particularly in the characterization of calcified lesions associated with atherosclerotic conditions. A theory of the field radiation patterns of disk and ring sources is presented which can be used to predict minima in the reflection spectra of a target as a function of frequency and angle relative to the acoustic axis. The theory is then tested experimentally for objects of known geometry by the use of a time delay spectroscopy system using a swept frequency technique, and results are obtained which demonstrate the usefulness of the technique. Finally, the theory is verified in the determination of lesion diameter and orientation from spectra received from calcified lesions on in vitro arterial specimens as a function of angle.
Image processing technique quantifies human atherosclerosis by computer analysis of arterial angiograms. X-ray film images are scanned and digitized, arterial shadow is tracked, and several quantitative measures of lumen irregularity are computed. In other tests, excellent agreement was found between computer evaluation of femoral angiograms on living subjects and evaluation by teams of trained angiographers.
A computer technique is being developed at the Jet Propulsion Laboratory to automate the measurement of coronary stenosis. A Vanguard 35mm film transport is optically coupled to a Spatial Data System vidicon/digitizer which in turn is controlled by a DEC PDP 11/55 computer. Programs have been developed to track the edges of the arterial shadow, to locate normal and atherosclerotic vessel sections and to measure percent stenosis. Multiple frame analysis techniques are being investigated that involve on the one hand, averaging stenosis measurements from adjacent frames, and on the other hand, averaging adjacent frame images directly and then measuring stenosis from the averaged image. For the latter case, geometric transformations are used to force registration of vessel images whose spatial orientation changes.
A computer system has been developed to quantify the degree of atherosclerosis in the human femoral artery. The analysis involves first scanning and digitizing angiographic film, then tracking the outline of the arterial image and finally computing the relative amount of roughness or irregularity in the vessel wall. The image processing system and method are described.
A computer image processing technique was developed to estimate the degree of atherosclerosis in the human femoral artery. With an angiographic film of the vessel as input, the computer was programmed to estimate vessel abnormality through a series of measurements, some derived primarily from the vessel edge information and others from optical density variations within the lumen shadow. These measurements were combined into an atherosclerosis index, which was found to correlate well with both visual and chemical estimates of atherosclerotic disease.
Femoral artery atheromatous lesions were studied and their changes as a measure of therapeutic effectiveness were assessed. The incidence of coronary risk factors in 100 patients was determined. Abnormal cholesterol was present in 42 percent, abnormal triglycerides in 66 percent, abnormal intravenous glucose tolerance test in 52 percent, judged from a K value of .9 or less by the technique of Wahlbert. A history of high blood pressure was present in 32 percent. Smoking was one of the most common factors.
The paper discusses the estimation of the degree of atherosclerosis in the human femoral artery through the use of a digital image processing system for vascular angiograms. The film digitizer uses an electronic image dissector camera to scan the angiogram and convert the recorded optical density information into a numerical format. Another processing step involves locating the vessel edges from the digital image. The computer has been programmed to estimate vessel abnormality through a series of measurements, some derived primarily from the vessel edge information and others from optical density variations within the lumen shadow. These measurements are combined into an atherosclerosis index, which is found in a post-mortem study to correlate well with both visual and chemical estimates of atherosclerotic disease.