The measurement of resolution in single photon emission computerised tomography.
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to S Webb.
Explore the source record for details and available documents.
Insulin-induced hypoglycaemia, which stimulates gastric acid secretion, is associated with an increase in circulating somatostatin levels in man. In order to assess the mechanisms involved in this rise, six normal volunteers connected to a Biostator for continuous glucose monitoring were studied, on three separate occasions. On each occasion after basal blood sampling, 0.15 i.u./kg body weight of insulin was administered i.v. and further samples were obtained intermittently over 150 min. On one occasion, dextrose was infused by the Biostator to prevent hypoglycaemia, while on the other two, a constant infusion of either normal saline or the specific H2 antagonist cimetidine was administered. Insulin plus dextrose caused no significant changes in circulating somatostatin levels, whereas insulin plus saline was associated with a marked, sustained and significant rise in all subjects; insulin plus cimetidine also produced a rise but it was delayed; the area under the curve was significantly (P less than 0.05) greater with insulin plus saline than with insulin plus cimetidine. These results show that in man insulin itself does not stimulate somatostatin secretion directly, but indirectly via hypoglycaemia. Further, the inhibition of gastric acid secretion with cimetidine reduces somatostatin release during insulin-induced hypoglycaemia. This suggests that gastric acid may mediate somatostatin secretion associated with insulin-induced hypoglycaemia.
A multiwire proportional-chamber positron camera, developed at the Rutherford Laboratory, has been evaluated at The Royal Marsden Hospital. The prototype camera consists of two opposing 30 X 30 cm2 chambers. Longitudinal tomograms of a positron-emitting radioactive distribution placed between the detectors are obtained via back-projection and 2D-deconvolution. Due to the limited stereoscopic angle achieved with stationary detectors, only five planes parallel to the detector faces are reconstructed. A selection of images is presented of phantoms using 68Ga and of patients using 18F, 18F fluorodeoxyglucose and Na124I, to illustrate the tomographic performance of the positron camera. A comparison between back-projected and deconvoluted images shows that the 2D-deconvolution process, which includes filtration of image noise, successfully removes the background due to scattered photons. The spatial resolution achieved depends on the half-height frequency cut-off used in the filtering process, and this parameter was chosen according to the count density in the back-projected images. A qualitative visual comparison was made between the positron images and equivalent single photon studies on the same patient. The results justify further development of this new detector, especially for use with generator-produced positron-emitting radionuclides. Three-dimensional deconvolution resulted in improved tomographic performance for phantom data, but was less successful for patient data. These problems associated with limited-angle tomography will be overcome by multi-view acquisition. This study has shown that a relatively low-cost positron imaging system can be used for routine organ imaging. Further developments in hardware and software should yield images which are superior to those from single-photon planar or tomographic studies.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Transverse section tomograms of experimental phantoms and patients have been obtained using a GE 400T camera and a filtered back-projection reconstruction technique. These tomograms have been compared with the corresponding sections reconstructed from the same tomographic projection data, but using iterative algorithms with correction for photon attenuation. The comparison assesses the importance of including a correction for attenuation as well as demonstrating how closely a simple geometric attenuation correction, applied to the filtered back-projection reconstruction method, approximates to a more accurate correction incorporated in the computation of line integrals during iterative reconstruction. A comparison is also made between the behaviour of reconstruction algorithms with simulated projection data and real data in terms of convergence properties, and some shortcomings arising from simulation are noted.
Explore the source record for details and available documents.
Single photon emission computed tomography (SPECT) of the liver and spleen has been studied using a GE400T/STAR rotating gamma camera/computer system. We have compared section image contrast, in both phantom and patient studies, produced by 180 degrees and 360 degrees reconstructions of data from a full 360 degrees data acquisition. There is little difference in overall image quality between the two reconstruction modalities. Quantitatively, however, appropriate 180 degrees reconstructions significantly improve lesion contrast in both organs by as much as a factor of two. In all cases studied, the values of image contrast achieved when the 180 degrees arc of reconstruction was centred about the region of interest were higher than the values obtained from the 360 degrees reconstructions. These results are shown to be independent of the corrections made for photon attenuation. Clinically this technique may prove useful in clarifying the presence or absence of disease, particularly in equivocal cases where lesions have poor contrast or are small in size. An extended clinical trial of the technique is now in progress.
An important feature determining the spatial resolution in transverse sections reconstructed by convolution and back-projection is the frequency filter corresponding to the convolution kernel. Equations have been derived giving the theoretical spatial resolution, for a perfect detector and noise-free data, using four filter functions. Experiments have shown that physical constraints will always limit the resolution that can be achieved with a given system. The experiments indicate that the region of the frequency spectrum between KN/2 and KN where KN is the Nyquist frequency does not contribute significantly to resolution. In order to investigate the physical effect of these filter functions, the spatial resolution of reconstructed images obtained with a GE 400T rotating gamma camera has been measured. The results obtained serve as an aid to choosing appropriate reconstruction filters for use with a rotating gamma camera system.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Exercise electrocardiography in women with chest pain is associated with a high incidence of false positive ST segment depression. The recent observation that changes in R wave amplitude during exercise can also be used diagnostically may improve the value of stress testing in women. The results of 12 lead treadmill exercise and coronary angiography were reviewed in 62 women, mean age 51 years, presenting with "angina" without previous myocardial infarction. These were compared with exercise results in 14 healthy asymptomatic volunteers with a mean age of 26 years. In addition to conventional ST analysis, R wave amplitude changes during exercise, measured in leads II, III, a VF, and V4 to 6, were examined. While the sensitivity and specificity of ST and R wave changes were similar at about 67%, their combined interpretation was helpful. If both ST and R wave criteria were negative the predictive accuracy for normal coronary angiography was 94% (17/18). Alternatively, in tests showing both ST depression and an abnormal R wave response, coronary angiography was always abnormal (13/13). None of the normal volunteers developed ST segment depression and 93% (13/14) had a normal R wave response. If both were positive, however, coronary angiography was always abnormal (13/13). Although stress test interpretation in women is difficult, R wave analysis is a useful adjunct to ST change and can improve the predictive accuracy of the test in a significant number of patients.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
A comparison is made between the performance of three transverse axial single-photon emission tomographic machines, namely a rotating gamma-camera and two different scanning systems. Each system has been evaluation in terms of the spatial resolution, sensitivity and efficiency. The functional dependence of the volume sensitivity on the size of the emitting object has been derived using a theoretical model of the photon emission, attenuation and detection. The model is shown to predict this size dependence well and is consistent with the experimental observations.
Explore the source record for details and available documents.
A Monte Carlo computer program has been used to calculate axial and off-axis depth dose distributions arising from the interaction of an external beam of 60Co radiation with a medium containing inhomogeneities. An approximation for applying the Monte Carlo data to the configuration where the lateral extent of the inhomogeneity is less than the beam area, is also presented. These new Monte Carlo techniques rely on integration over the dose distributions from constituent sub-beams of small area and the accuracy of the method is thus independent of beam size. The power law correction equation (Batho equation) describing the dose distribution in the presence of tissue inhomogeneities is derived in its most general form. By comparison with Monte Carlo reference data, the equation is validated for routine patient dosimetry. It is explained why the Monte Carlo data may be regarded as a fundamental reference point in performing these tests of the extension to the Batho equation. Other analytic correction techniques, e.g. the equivalent radiological path method, are shown to be less accurate. The application of the generalised power law equation in conjunction with CT scanner data is discussed. For ease of presentation, the details of the Monte Carlo techniques and the analytic formula have been separated into appendices.