Ernest V. Garcia

Analysis of the degree of pulmonary thallium washout after exercise in patients with coronary artery disease

An abnormal increase in pulmonary thallium activity may be visualized on post-stress thallium images in patients with coronary artery disease. Because this increased pulmonary thallium activity usually disappears by the time of redistribution imaging, this study was designed to assess whether measurement of the degree of pulmonary thallium washout between stress and redistribution might improve the detection of increased pulmonary thallium activity in patients with coronary artery disease. Quantitative analysis revealed abnormal (that is, greater than 2 standard deviations of normal values) pulmonary thallium washouts in 59 (64%) of 92 patients with coronary artery disease, but in only 2 (25%) of 8 subjects with angiographically normal arteries (p less than 0.06). By comparison, the visual analysis of pulmonary thallium washout and use of initial pulmonary to myocardial thallium ratio were significantly (p less than 0.05) less sensitive in detecting abnormality in patients with coronary artery disease. Abnormal pulmonary thallium washout was related to both the anatomic extent and functional severity of disease: it occurred with greatest frequency in patients with multivessel disease and in those with exercise-induced left ventricular dysfunction (p less than 0.005). When added to the quantitative analysis of myocardial scintigraphy, the analysis of pulmonary thallium washout increased the detection of coronary artery disease from 84 to 93% (p less than 0.05), but the sample size was too small to assess specificity. Thus, the analysis of pulmonary thallium washout is a useful diagnostic variable because it: 1) provides an objective measurement of abnormal pulmonary thallium activity and is more sensitive than other methods; 2) correlates with both the extent of coronary artery disease and the degree of exercise-induced left ventricular dysfunction, and 3) improves the sensitivity of quantitative myocardial thallium scintigraphy to detect the presence of coronary artery disease.

Digital processing in cardiac imaging

SummaryIn recent years, imaging modalities have realized the potential for digitizing images of the heart. These modalities include nuclear cardiology, echocardiography, digital subtraction angiography, computed tomography, positron emission tomography, and magnetic resonance imaging. Once the cardiac images have been digitized by the computer and formatted in a pixel array of data, the computer can perform mathematical operations on the input (original) images for the purpose of providing improved output (processed) images. The computer can also analyze the original (or processed) images for the purpose of extracting global, regional, or temporal measurements of cardiac perfusion and function. These processes include the restoration, enhancement, analysis, manipulation, and coding of images. These basic processes are implemented by a combination of image processing operations. These operations include pixel point processing, pixel group processing, frame processing, geometric processing, and information extraction. This article describes how these operations are used to perform processes, and how these processes are applied to cardiac imaging, currently and in the future.

Quantitative analysis of the distribution and washout of thallium-201 in the myocardium: description of the method and its clinical applications

Exercise thallium-201 (Tl-201) myocardial perfusion scintigraphy is a useful noninvasive method for detecting and evaluating patients with significant coronary artery disease (CAD). Visual interpretation of analogue Tl-201 images for evaluation of regional myocardial perfusion, even when performed by experienced readers, is subject to substantial observer variability [1]. This approach is further limited by dependence on the quality of the hard copy output and the inability to accurately compensate for background activity. Furthermore, although the visual method is highly specific for localization of CAD, it has major limitations in sensitivity for detection of individual coronary stenoses [2, 3], especially in patients with multiple vessel CAD [2–6] and when the degree of coronary stenosis is not severe [2,3]. Finally, although the regional washout characteristics of the myocardium for Tl-201 contain important diagnostic information, washout abnormalities are difficult to detect by visual inspection.

Myocardial perfusion imaging by single photon emission computed tomography (SPECT)

Single Photon Emission Computed Tomography (SPECT) is superior to the planar imaging method because it provides a higher image contrast resolution and allows separation of overlapping myocardial regions. Several computerized methods are now available and are being developed for quantitative analysis of SPECT myocardial perfusion by Tl-201 or the new Tc-99m labeled myocardial perfusion agents. In patient studies, all short axis and apical portions of vertical long axis TI-201 SPECT images are quantified by dividing each myocardial slice into 60 equal sectors and displaying the maximal count per sector as a linear profile. The best threshold for defining normal limits was developed in a pilot group of 45 patients. After comparing patients’ profiles with normal limits, abnormal and normal portions of the patients’ profiles are plotted on a 2-dimensional polar map which is divided into specific coronary artery territories based on a scheme developed in a group of patients with disease of different coronary arteries. ROC analysis for defect size showed that the optimal thresholds for a definite perfusion defect were 12 % for the LAD and LCX and 8 % for the RCA territories. These criteria were prospectively applied to an additional 138 patients which yielded respective sensitivity, specificity and normalcy rates for overall detection of CAD of 96 %, 56 % and 86 % with high sensitivity and specificity for identification of CAD in individual coronary arteries. The accuracy of this quantitative SPECT technique was further assessed in a multicenter trial consisting of 318 patients whose SPECT images were obtained by various cameras, computers and operators. The results indicated that the quantitative SPECT method, utilizing standard normal limits developed at Cedars-Sinai Medical Center, can be applied at other institutions with similar accuracies. In 66 patients with prior myocardial infarction, new quantitative criteria were developed by ROC analysis that took into consideration contiguity of defects with the infarct zone. The new defect thresholds (40 % for LCX, 20 % for RCA and 12 % for LAD) were 86 % accurate for detection of patients with multivessel coronary disease after myocardial infarction. SPECT is superior to the planar imaging method in detecting patients without prior myocardial infarction, and those with moderate or single vessel coronary disease. SPECT is increasingly being used in conjunction with Tc-99m labeled myocardial perfusion agents. The results of qualitative analysis of SPECT Tc-99m SestaMIBI studies in a multicenter study have been similar to those of TI-201 SPECT for detection of perfusion defects and evaluation of the patterns of defect reversibility. Using an approach similar to that used for quantitation of TI-201 SPECT studies, a quantitative method has recently been developed for the interpretation of exercise-rest Tc-99m SestaMIBI images. Furthermore, methods are being developed for the analysis of same day rest-stress protocols and for the absolute quantification of myocardial perfusion by performing attenuation and scatter correction on Tc-99m SestaMIBI myocardial perfusion images. Myocardial perfusion SPECT images are being quantified with respect to the extent of myocardial perfusion deficit which holds promise for assessing percent infarcted and jeopardized myocardium as important prognostic indicators in coronary artery disease.

Single Photon Emission Computed Body Tomography Using a Multiplane Scanner

This paper addresses a computer interface and required software developed for data acquisition and tomographic reconstruction using a standard nuclear medicine computer (32K Modumed-Medical Data Systems) interfaced to the Anger Multiplane Tomographic Scanner (PhoCon-Searle Radiographics). This combination provides single-photon emission computed body tomography, with longitudinal and transverse section capabilities. During acquisition the computer digitizes and records on magnetic disk the information generated by each scintillation event. Serial mode acquisition is utilized to create a list of the coordinates of the scintillation events in the crystal and which probe detected the event. This list contains markers every 10 milliseconds. These markers are pointers to the X coordinate of the probes. The Y coordinate of the probes are stored in a directory at the first disk record. Tomograms are reconstructed by back-projection of each scintillation event along a line passing through the focal point of the collimators. Programs have been developed to reconstruct one or multiple longitudinal or transverse sections. Once the image has been reconstructed, standard image manipulations such as contrast enhancement, color display, and movie display can be performed.