Search PubMed⌕ Search

Biomedical subjects

A S Houston

Publications and source records attributed to A S Houston.

At least 37 records · Page 2Linked to original sources
PubMed1989

Factor analysis of dynamic structures (FADS) in the diagnosis of renal disease.

Factor analysis of dynamic structures (FADS) has been used in the interpretation of dynamic scintigraphic studies since the technique was described by Bazin et al. (1980). This study was designed to analyse to what extent, if any, does physiological factor analysis of dynamic renal data really help the clinician and by how much the method improves the diagnostic accuracy when compared to deconvolution analysis and parenchymal transit time (PTT) measurements. One hundred and fifty patients who were clinically, biochemically and radiologically investigated for renal disease were included in the study. Fifty of these were found to have no clinical evidence of renal disease, 50 were diagnosed as having non obstructive kidney disease and 50 had evidence of renal obstruction. Data obtained from 99mTc-DTPA renography were processed using deconvolution (with PTTs) and physiological factor analysis and the results compared by ROC analysis. Clinically the information gained from factor analysis was superior to that obtained from deconvolution with PTT measurements in that a more accurate differentiation between an obstructive nephropathy and an obstructive uropathy was obtained. It is considered that physiological factor analysis enhanced the clinical information obtained from renography, increases diagnostic accuracy and obviates the need for diuresis renography.

Humans↗
PubMed1989

A perception experiment to assess the clinical accuracy of various methods of obtaining cardiac phase images.

Five methods of obtaining cardiac phase (and phase-like) images are compared. These are single-harmonic Fourier phase imaging; time-to-minimum using a two-harmonic fit to each dixel; time-to-minimum using four weighted harmonics; composite factorial phase imaging; and individual factorial phase imaging. Polaroid prints of phase images for both gated equilibrium studies (using all methods) and first-pass studies (first and last methods only) were shown to observers who were asked to rate the images according to their confidence of an abnormality presenting. A receiver operating characteristic (ROC) analysis was performed. For gated studies composite factorial phase imaging appears to be best, while all methods are significantly better than single-harmonic Fourier phase imaging. For first-pass studies individual factorial phase imaging appears to be superior to single-harmonic Fourier phase imaging.

Heart↗
PubMed1989

Comparison of two interpolative background subtraction methods using phantom and clinical data.

Two interpolative background subtraction methods used in scintigraphy are tested using both phantom and clinical data. Cauchy integral subtraction was found to be relatively free of artefacts but required more computing time than bilinear interpolation. Both methods may be used with reasonable confidence for the quantification of relative measurements such as left ventricular ejection fraction and myocardial perfusion index but should be avoided if at all possible in the quantification of absolute measurements such as glomerular filtration rate.

Glomerular Filtration Rate↗
PubMed1984

The effect of apex-finding errors on factor images obtained from factor analysis and oblique transformation.

The use of an oblique transformation to obtain physiological factors from a factor analysis of dynamic studies has recently been suggested. The technique involves using an apex-seeking routine to enable these factors and their contribution images to be formed. This paper investigates the effects of errors in apex-finding on these factor images. In particular it is shown that, apart from scaling, the factor image obtained from a given apex is independent of the position of that apex (and hence is independent of its own factor).

Humans↗
PubMed1982

Factorial phase imaging: a new concept in the analysis of first-pass cardiac studies.

A method is described which offers an alternative to conventional Fourier phase/amplitude imaging for analysing representative LV cycles obtained from first-pass cardiac studies. The method involves applying factor analysis to the data and producing functional images of the first two factor loadings after polar transformation. The method corrects for cycle asymmetries which produce anomalies in the Fourier images. For 88 studies the method provides greater uniformity among normal phase images while retaining, or even enhancing, abnormalities. Small differences between amplitude images produced by the two methods do not appear to be important.

Fourier Analysis↗
PubMed1982

Functional imaging in the early diagnosis of dysbaric osteonecrosis.

Eight divers who developed permanent scintigraphic bone changes following dives to various depths took part in a study to determine the nature of these lesions. Only one of these divers had radiologically evident bone changes in his scintigraphic lesion. Dynamic scintigrams of the bones containing the lesions were obtained from all eight divers and functional images generated of both amplitude (osteoblastic activity) and accretion (microvasculature) rate constants. All had increased amplitude in the area of the scintigraphic lesion indicating continuing osteoblastic activity. Three divers showed a decreased accretion rate in the lesions, indicating impairment or absence of local microcirculation; one of these was the diver who showed radiological changes prior to the study and the other two subsequently developed radiologically evident osteonecrosis. It is concluded that functional images, generated from dynamic bone scintigrams, can successfully predict dysbaric osteonecrosis.

Barotrauma↗
PubMed1981

A comparison of three methods of assessing renal function.

In a study of 100 patients referred for assessment of renal function a comparison is made between gamma camera computer assisted displays of 99mTc-DTPA(Sn) nephrograms, intra-venous urography (I.V.U.) and triple probe 131I-Hippuran renography. The computer processed data displays renal morphology and activity/time curves which are deconvolved enabling quantitative assessment of glomerular function to be made. Pattern recognition techniques for feature extraction are employed to facilitate classification of the curves. It is concluded that the computer processed data gives better results both in the recognition of morphological defects and in the indication and measurement of renal damage.

Glomerulonephritis↗
PubMed1980

Processing of liver dynamic studies with technetium-labelled sulphur colloid.

The uptake of technetium-labelled sulphur colloid by the liver was investigated by monitoring 69 patients on a gamma camera/computer system. Following an intravenous injection 30 half-minute frames were collected and summed to give a composite view. Regions of interest were drawn round the liver and over the aorta and corresponding activity-time curves were formed. A simple physiological model has been assumed and two parameters, namely total uptake and rate of uptake, were shown to give fairly good separation between the classes of normal, cirrhosis, space occupying lesions, reticulosis and hepatitis. Furthermore the uptake rate-constant was also shown to be a useful indication of whether or not a reticulosis or hepatitis was in an acute phase. Principal components analysis was also applied to the liver curves producing two relevant components. However, a plot of the first against the second yielded no better separation of classes than that obtained by the model.

Colloids↗
PubMed1980

A comparison of four standard scintigraphic TV displays: concise communication.

Four displays (pseudocolor mappings) available on a standard color TV are compared using two series of images (100 normal brains and 100 normal livers) with 50 computer-simulated lesions superimposed on each set. Four observers viewed the sets of images in such a way that the order of the display methods for both organs formed two orthogonal Latin squares. The observers were asked to locate and rate, on a standard scale, the most apparent area in each image, and ROC analysis was applied to the results. The "heated-object spectrum" was shown to be a useful display for brain images. It was also shown that the choice of display depends on the organ to be imaged.

Brain↗
PubMed1979

Principal components analysis as an aid to classification of renal dynamic studies.

Four hundred renal dynamic studies obtained using 99mTc-(Sn) DTPA were classified clinically into four classes (normal, pre-renal lesion, intrarenal lesion and urinary tract obstruction). Principal components analysis was then applied to the kidney activity/time curves and yielded good class separation using only three components. The separation of normal and obstructed or damaged kidneys using only the first component was better than that obtained using the calculated mean transit time.

Diagnosis, Differential↗