Search

Twenty patients who had undergone a heterotopic heart transplant were studied prospectively to determine the relationship between rejection and ventricular dysfunction assessed from gated blood-pool studies. A fully automated method for detecting ventricular edges was implemented; its success rate for the grafted left and right ventricles was 94% and 77%, respectively. The parameters, peak ejection and filling rates, were calculated pixel per pixel using a two-harmonic Fourier algorithm and then averaged over the ventricular region of interest. Peak filling and ejection rates were closely related with the severity of the rejection, while the left ventricular ejection fraction was not. Peak filling rates of both ventricles were the indices closely related to the presence of moderate rejection. Despite the low number of patients, these data suggested that gated blood-pool-derived indices of ventricular function are associated with ventricular dysfunction resulting from myocarditis rejection. Radionuclide ventriculography provides parametric data which are accurate and reliable for the diagnosis of rejection.

A semi-automatic method was developed to determine the anterior (AC) and posterior (PC) commissures on brain single-photon emission tomographic (SPET) perfusion images, and then to draw the proportional anatomical Talairach's grid on each axial SPET image. First, the AC-PC line was defined on SPET images from the linear regression of four internal landmarks (frontal pole of the brain, inferior limit of the anterior corpus callosum, sub-thalamic point and occipital pole). Second, the SPET position of AC and PC points on the AC-PC line was automatically determined from measurements made on hard copies of magnetic resonance (MR) images of the patients. Finally, a proportional Talairach's grid was automatically drawn on each axial SPET image. To assess the accuracy of localization of AC and PC points, co-registered technetium-99m hexamethylpropylene amine oxime SPET and MR images from 11 subjects were used. The mean displacements between estimated points on SPET and true points on MRI (?x=sagittal, ?y=frontal and ?z=axial displacement) were calculated. The mean displacements (in mm) were ?x=-1.4±1.8, ?y=-1.7±3.3 and ?z=-1.1±2.5 for AC, and ?x=-1.8±1.8, ?y=0.3±3.2 and ?=-1.3±2.7 for PC. These displacements represented an error of less than 5 mm at the anterior or posterior pole of the brain or at the vertex. Intra- and inter-observer comparisons did not reveal significant differences in mean displacements. Thus, this semi-automatic method results in reproducible and accurate stereotactic localization of SPET perfusion abnormalities. This method can be used routinely for repeat follow-up studies in the same subject as well as in different individuals without requiring SPET MRI co-registration.

A data base of clinical studies is required for quality assurance of software used for analysis of radionuclide cardiac imaging procedures. Studies used must be rigorously validated in terms of both the clinical condition of the patient undergoing the procedure and the imaging protocol used. Selection protocols for the creation of a software phantom data base of normal studies and three typical patterns of cardiac disease — recent transmural myocardial infarction, isolated myocardial ischaemia and dilated cardiomyopathy — have been developed by the Cardiac Working Group of the European COST B2 project in association with the Cardiac Task Group of the European Association of Nuclear Medicine and the Working Group on Nuclear Cardiology of the European Society of Cardiology. These protocols include criteria for the interpretation of qualitative and quantitative non-radionuclide data. Compliance of the clinical data with the selection criteria will have to pass scrutiny by an international team for each study used as a software phantom. The radionuclide studies encompass stress and rest myocardial perfusion studies (planar and single photon emission tomography) using thallium-201 and technetium-99m methoxyisobutylisonitrile and rest gated blood pool studies. Methods for acquisition of data are defined for each type of study and for each individual study a portfolio of all clinical data is established. A pilot study is required to investigate the problems and logistics of distributing clinical radionuclide studies between a range of computers and institutes, and to ascertain the procedures necessary for analytical comparison of the results obtained.

Abstract: ECG gated201T1 scintigraphy was compared to non gated images for detecting 2–3 vessel disease 1–3 months after a first uncomplicated myocardial infarction. In all, 111 patients on anti anginal treatment underwent coronary arteriography and stress thallium imaging; 39 showed single vessel disease (SVD), and 72 multivessel disease (MVD). Sensitivity of black and white analog images was 82% for SVD and 8% for MVD. Sensitivity of computerized colored static images was 87% for SVD and 33% for MVD. For gated images, sensitivity was 100% and 92% in patients with SVD and MVD respectively. Specificity for detecting MVD was 95% for black and white images, 77% for computerized colored static images, and 69% for ECG gated images. Thus, ECG gated exercise201TL scintigraphy is useful for predicting the extent of coronary artery disease 1-3 months after myocardial infarction in patients on anti anginal therapy.

Stress 201Tl myocardial images were obtained using a standard ECG gated acquisition protocol for 10 min per view. These images were filtered using a three dimensional space and time filter. The effects of filtering were evaluated by sectorial analysis. Comparing raw and filtered dynamic images, quantification demonstrated that filtering did not produce artefacts. A conventional static image was also obtained as the sum of the series of unfiltered dynamic images. Comparison of raw static and filtered diastolic images indicated that the latter demonstrated myocardial abnormalities more clearly (P<0.01) in 33 patients. Improvement in the signal to noise ratio produced by filtering made cinematic display and visual analysis possible which was not feasible with raw gated images. Clinical evaluation is in progress.

An algorithm has been written which automatically selects a left epicardial region of interest on 201Tl myocardial images. It accomplishes a radial search from the geometric center of the myocardium. On each of the 30 profiles, the local maximum of the 1st derivative is selected as the epicardial edge. The algorithm has been tested in 40 patients at stress and redistribution. In ten patients, conventional planar images were obtained in three views. In 30 patients conventional short axis tomographic images were obtained after reconstruction of 32 projections acquired over 180°.

A fully automated data acquisition, processing, and display procedure was developed for equilibrium radioventriculography. After a standardized acquisition, the study is automatically analzyed to yield both right and left ventricular time-activity curves. The program first creates a series of edge-enhanced images (difference between squared images and scaled original images). A marker point within each ventricle is then identified as that pixel with maximum counts to the patient's right and left of the count center of gravity of a stroke volume image. Regions of interest are selected on each frame as the first contour of local maxima of the two-dimensional second derivative (pseudo-Laplacian) which encloses the appropriate marker point, using a method developed by Goris. After shifting the left ventricular end-systolic region of interest four pixels to the patient's left, a background region of interest is generated as the crescent-shaped area of the shifted region of interest not intersected by the end systolic region. The average counts/pixel in this background region in the end systolic frame of the origina study are subtracted from each pixel in all frames of the gated study. Right and left ventricular time-activity curves are then obtained by applying each region of interest to its corresponding background-subtracted frame, and the ejection fraction, end diastolic, end systolic, and stroke counts determined for both ventricles. In fourteen consecutive patients, in addition to the automatic ejection fractions, manually drawn regions of interest were used to obtain ejection fractions for both ventricles. The manual regions of interest were drawn twice, and the average obtained. For the right ventricle, the correlation between auto and average manual ejection fraction was 0.52; the correlation between the two manual ejection fractions was 0.88. For the left ventricle, the correlation between auto and average manual ejection fraction was 0.96; the correlation between the two manual ejection fractions was 0.91. Automated processing is essential for the accurate and reproducible assessment of left ventricular ejection fraction.

We have developed a simple method for measuring left ventricular volume based on semi-automated analysis of 40° left anterior oblique images obtained with a standard scintillation camera after equilibrium of an intravenous injection of 20 mCi of technetium-99m in vivo labeled red blood cells. The essence of the method is the use of the dimensions and radioactivity within a segment of aorta to convert observed left ventricular count rates to volume. Four assumptions were made: 1) the aortic arch is nearly parallel to the collimator face when a patient is in the proper left anterior oblique position; 2) a segment at the top of the aortic arch, approximately 1 cm wide, is a right cylinder, 3) the edges of the aorta can be delineated as the lines where the second derivative of a cross sectional profile equals zero; 4) left ventricular and aortic arch counts undergo the same attenuation because they are nearly the same distance from the chest wall in the proper left anterior oblique position. By measuring the counts and volumes of two regions of known shape, one in the middle, the other at the edge of the aortic arch, and calculating their differences a background-independent volume count ratio (?v/?C) can be obtained. The left ventricular and diastolic volume (LVEDV) is calculated with the equation: LVEDV=(?/?C) LVEDC, where LVEDC represents left ventricular end diastolic counts. Twenty-six patients were evaluated by equilibrium radio- and contrast-ventriculography, the latter analyzed by planimetry. The radionuclide method yielded an end diastolic volume that correlated well with contrast ventriculography (r=0.96, Y=0.91 X+21 ml). In addition to its simplicity and objectivity, a major advantage of this method of determining ventricular volume is that it does not require a blood sample.