R. A. Shields

Attenuation correction of myocardial Spect studies using low resolution computed tomography images

ObjectivesArtefacts caused by tissue attenuation can create problems in the interpretation of myocardial perfusion studies. This study aimed to determine if attenuation correction using the ‘Hawkeye’ low-resolution X-ray computed tomography facility attached to a GE Millennium VG gamma camera reduced attenuation artefacts in our patient group. This technique offers potential advantages over isotope transmission methods of attenuation correction because of the lack of noise in the transmission images. MethodsNinety-four patients (64 males, 30 females) underwent stress/rest myocardial perfusion imaging using 99mTc tetrofosmin (188 studies). Semi-quantitative analysis of the attenuation corrected and non-corrected perfusion images was carried out using proprietary software (Emory ECToolbox). ResultsNo significant differences between attenuation corrected and non-corrected images were seen in the septal, lateral or apical segments. Attenuation correction produced a significant reduction in the defect scores in the inferior segment indicating the presence of attenuation artefacts. This was more marked in male patients. The incidence of attenuation artefacts in the inferior segment increased with weight although patients of normal weight also showed attenuation artefacts. The changes in the anterior segment were more difficult to interpret particularly in females, with attenuation correction increasing the defect score in some patients. Minor degrees of mal-registration (even of 1 pixel) between emission and transmission images increased the likelihood of creating a defect. ConclusionAttenuation correction using Hawkeye is likely to improve diagnostic accuracy in men, but is less likely to be useful in women.

Technegas and pertechnegas particle size distribution

Technegas and Pertechnegas are radioactive aerosols produced in a commercial generator and used for lung scintigraphy. The aerosols are produced by first evaporating to dryness standard technetium-99m generator eluate (99m-TcO4 in normal saline) in a graphite crucible (thesimmer stage) and then heating this to 2500° C (the „burn” stage). The aim of this work was to measure the particle size distributions of these agents and relate this to regional lung deposition. Factors that may vary during production of the aerosol in routine use were investigated to determine whether they influenced the particle size. Activity size distributions were measured using a serial wire-screen diffusion battery. The Technegas size distribution was approximately log-normal with an activity median diameter of 158 nm and a geometric standard deviation of 1.5. The results for Pertechnegas were similar. The median size increased with the number of simmers and with the time from generation. The increase in size with the number of simmers is thought to be due to the increased salt content in the crucible prior to the „burn”. The predicted lung deposition is 37% in the alveolar region and 5% in the bronchial region. Significant changes in deposition are not predicted over the range of particle sizes measured.

The plasma protein binding of HIDA.

By using Sephadex gel column chromatography to separate substances into their various components according to molecular weight, we have investigated the effect of incubating several “brands” of HIDA in plasma, in vitro.The results show that such incubation has no effect on either dimethyl HIDA, or diethyl HIDA, but that in the case of para-butyl HIDA, incubation in plasma increases the Rf value to that of HSA (human serum albumin).This indicates that para-butyl HIDA becomes bound to plasma proteins, in contrast to both dimethyl HIDA and diethyl HIDA.

The use of the L-curve method in the inversion of diffusion battery data

Regularisation techniques are often employed to find a unique solutions to the diffusion battery data inversion problem. This involves minimising a function of the form A + λB, where A measures the fit to the data and B measures the smoothness of the solution. The quality of the solution is critically dependent on the choice of λ. if A is plotted against B for all λ values then a curve with a characteristics L-shape is obtained and the solution at the corner has an optimum balance of fit and smoothness. We demonstrate the usefulness of the L-curve method in this application and show that it provides a good means of choosing λ. It has the particular advantage that solutions are independent of error estimates. We also show that the solutions obtained are superior to those obtained by Twomeys non-linear inversion algorithm.

Diuresis renography and parenchymal transit times in the assessment of renal pelvic dilatation

A comparison has been made between the results of diuresis renography and deconvolution analysis of the renogram in 36 patients (46 kidneys) with urographically demonstrated renal pelvic dilatation. The mean transit times obtained by deconvolution analysis of the parenchymal renograms of eight normal volunteers had a mean value of 2.8 + 0.6 (S.D.) min, and the upper limit of normal was thus established as 4 min. Twenty-two of 29 kidneys (76%) with nonobstructive diuresis renograms had parenchymal mean transit times of 4 min or below. Fourteen out of 17 kidneys (82%) with obstructive diuresis renograms had parenchymal mean transit times above 4 min. Both investigations may contribute to the noninvasive assessment of dilated renal pelves; the parenchymal retention function (quantified by mean transit time) assesses the effects of dilatation on tubular transit and renal function, whereas the diuresis renogram is a more specific test for evaluating impedance to flow.

A rapid quantitative method for the preparation of 123I-iodo-hippuric acid

The labelling of o-iodo-hippuric acid (hippuran) with 123I by several methods was investigated with a view to producing a simple one-step kit preparation. The requirements for the final product are high labelling efficiency and high radiochemical purity. A method which incorporates CuSO4·5H2O as a catalyst was found which gave promising preliminary results. Such variables as pH, mass of CuSO4·5H2O, volume of 0.02 N NaOH containing the 123I, reaction temperature and time were investigated in detail and optimum values obtained. The results have led to the production of a kit which can be stored for up to 3 months in liquid form and up to 9 months if freeze-dried. The kit has been used in several thousand patients and has yielded consistently good clinical results.

The preparation and composition of HIDA

The compositions of four “brands” of Tc-HIDA — dimethyl HIDA (Sorin Biomedica), diethyl HIDA (Solco), diethyl HIDA (The Radiochemical Centre), and para-butyl HIDA (Sorin Biomedica) —were investigated at varying times after their preparation from “instant kits”, by using Sephadex gel column chromatography.Up to three peaks of activity were seen in each chromatogram, representing free reduced technetium, the monocomplex, and the biscomplex of Tc-HIDA respectively.In all cases the free reduced technetium component diminished to reach a plateau at less than 4% of total activity, by 1h. In both “brands” of diethyl HIDA there was a transformation from the monocomplex to the biscomplex which took about one hour. The rate of this transformation was increased by raising the temperature of incubation, or by agitation during incubation.In the para-butyl HIDA this transformation was again noted, but was incomplete. In the dimethyl HIDA the monocomplex form was not seen.There was no noticeable disintegration of any Tc-HIDA complex up to 6 h.It is suggested that HIDA should be incubated for 1 h after preparation, before use, to achieve a stable composition.

Comparison of local and global optimisation techniques for diffusion battery data analysis

The recovery of a size distribution from a set of diffusion battery deposition measurements is a typical example of a data inversion problem. A range of solution methods have been proposed and most rely on an iterative optimisation procedure. We used a non-linear regularisation approach and investigated the differences between local and global optimisation techniques. For local search methods the final solution depends on the starting point of the search and the probability of finding a good solution decreases as the search space becomes more complex. We present a global optimisation method based on simulated annealing. It is shown that good solutions can be consistently found using this method, but, that considerable computing time is required. The same solutions were found far more rapidly by performing repeats of a local search method from a range of start points. We conclude that the simulated annealing technique offers little advantage in the present application, although it has some attractive theoretical and practical properties and may prove to be useful in other aerosol data inversion problems.

Study of right ventricular function in ischaemic heart disease using radionuclide angiocardiography

A technique for the estimation of LVEF from first passage radionuclide angiocardiography was adapted to provide estimates of RVEF. In 17 subjects with no history of cardiovascular disease mean LVEF was 0.71±0.08 and mean RVEF 0.65±0.08. Mean values for 15 subjects with coronary artery disease but no previous history of myocardial infarction were 0.66±0.10 for LVEF and 0.65±0.08 for RVEF. Depressed ejection fractions were found after acute myocardial infarction. LVEF was lower after anterior (0.43±0.06) than inferior (0.51±0.10) infarction. RVEF was normal in the majority of subjects with anterior infarction (0.58±0.10) but was depressed after inferior infarction (0.50±0.05). Similar, although less marked, results were found in a group of subjects with old myocardial infarction.

The influence of inhalation technique on Technegas particle deposition and image appearance in normal volunteers

The aim of this work was to investigate the influence of inhalation technique on Technegas image quality and on fractional particle deposition. This was investigated in six normal volunteers using three different types of breathing pattern. Fractional deposition was determined by analysis of dynamic gamma camera images acquired during Technegas administration. Static lung images were subsequently acquired and assessed independently by three experienced observers. High-quality images were obtained in all cases although slight differences were noted. The images produced using a slow deep inspiration with a breath hold (i.e. the standard method) were of more uniform texture and also had the least gradient in activity from apex to base. The converse was true for a rapid inhalation technique. The average fractional deposition per breath was 55%, but this varied between individuals and with breathing pattern, being most influenced by the total duration of a breath. We conclude that for patient studies the standard inhalation technique is best, although variation to suit individual patients would be acceptable. These results contrast with similar studies using conventional radio-aerosols, which tend to show a greater sensitivity of image appearance to changes in breathing pattern and lower fractional depositions.