Margaret A. Simons

The influence of expertise on X-ray image processing.

Observers with four different levels of radiological experience performed a recognition memory task on slides of faces and chest X-ray films. Half of the X-ray films revealed clinically significant abnormalities and half did not. Recognition memory for faces was uniformly high across all levels of radiological experience. Memory for abnormal X-ray films increased with radiological experience and, for the most experienced radiologists, was equivalent to memory for faces. Surprisingly, recognition memory for normal films actually decreased with radiological experience from above chance to a chance level. These results indicate that radiological expertise is associated with selective processing of clinically relevant abnormalities in X-ray images. Expert radiologists appear to process X-ray images the way that we all process faces, by quickly detecting and devoting processing resources to features that distinguish one stimulus from another. However, the selective processing of X-ray films appears to be restricted to clinically relevant abnormalities. As they develop the ability to detect these abnormalities, radiologists appear to lose the ability to detect variations in normal features.

A mathematical model for signal from spins flowing during the application of spin echo pulse sequences

Models are presented for both laminar and plug flow that predict the signal from spins flowing during the application of slice-selective spin echo pulse sequences. The models permit calculation of the total signal from a cylindrical vessel lying perpendicular to the slice and incorporate the effect of the physical displacement of the spins between successive excitations. This time-of-flight effect gives a signal which is composed of contributions from a finite number of spin populations, with each population signal weighted by the fractional volume of that spin population within the cylindrical vessel segment. The signal and fractional volume from each spin population are derived analytically for ten different spin echo pulse sequences. The models for plug and laminar flow have important application for predicting and interpreting flow effects observed in clinical images. They are shown to be useful for selecting pairs of pulse sequences that can be used to obtain digitally subtracted MR images which provide optimum contrast for flowing blood with essentially complete suppression of stationary anatomy. These models provide a means for quantitatively comparing the expected signal from flowing spins for the many techniques presently being investigated for MR angiography.

Vessel Diameter Measurement Using Digital Subtraction Radiography

A method for obtaining absolute diameter and cross-sectional area measurements on subtraction digital images is described and tested in phantom vessels from 1.5 to 5.5 mm in diameter filled with iodine contrast at concentrations from 23 to 185 mg I/ml. A highly linear correlation of true vs. calculated diameter is demonstrated, with accuracy and reproducibility of the method varying from +/- 1% to 2% at the highest iodine concentration to +/- 30% in the smallest tube at the lowest concentration. A method is described for correction of the observed video density values to allow for nonlinearity of response of the imaging system to iodine density, and its effect on the measured diameters is demonstrated.

Effects of Carbenoxolone on Gastric Mucosal Permeability and Blood Flow in the Dog

The effects of topical application of carbenoxolone at neutral and acidic pH were compared in exteriorized, chambered segments of canine gastric corpus. When dissolved in saline at pH 7.5 to 8.0, 0.25% carbenoxolone caused a rapid drop in gastric potential difference of 56 +/- 2 mv and greatly increased permeability to H+ ions. Blood flow, as measured by radioactive microspheres, was not changed by carbenoxolone treatment, but subsequent exposure to isotonic HC1 caused an abrupt rise in flow. Application of 0.25% carbenoxolone suspension in isotonic HC1 caused no change in potential difference, permeability, or blood flow. Neither carbenoxolone preparation had a significant effect on aspirin-induced H+ back-diffusion or injury.

Quantitative Digital Subtraction Coronary Angiography Using Videodensitometry An In Vivo Analysis

A videodensitometric method for measuring absolute cross-sectional area and diameter has been tested in living dogs with coronary artery stenoses created surgically by placement of small Silastic cuffs. Coronary arteriograms were performed using a circular tomographic unit to provide multiple views of each lesion, and measurements were made from logarithmically subtracted digital images. Dimensions of 13 stenoses of cross section 1 to 5 mm2 and adjacent reference segments (2 to 9 mm2) were determined by histologic sectioning of the segments after injection with a rapidly hardening plastic fixative under physiologic pressure. Two different methods were tested for calculating cross-sectional area. On 238 measurements, 102 of normal vessel segments and 135 of stenoses, both methods showed good correlation with histologic measurements, with slopes of 0.929 + (SD) 0.037 (r = 0.8563) and 0.948 + (SD) 0.037 (r = 0.8554). Multiple measurements of each segment produced values within 30% of the true absolute cross sectional area in most cases. The method shows promise as a means for quantitating absolute dimensions of vessels in clinical arteriography.

Comparison of observer and videodensitometric measurements of simulated coronary artery stenoses

To test the applicability of an automated vessel measurement technique to coronary arteriography, a videodensitometric method with caliper measurements on digital subtraction images of a moving coronary artery phantom was compared. Percent diameter stenosis was determined by both methods, revealing a twofold improvement in reproducibility with the videodensitometric method, with percent stenosis being determined within +/- 10% for two different iodine concentrations injected during continuous flow into the simulated coronary arteries. Absolute diameters were also measured by the videodensitometric method, showing a high degree of correlation between measured and true diameter for vessels between 0.5-3.0 mm.

Clinical evaluation of a 2K x 2K workstation for primary diagnosis in pediatric radiology

Preliminary results of a large-scale ROC study evaluating the diagnostic performance of digital hardcopy film and 2K X 2K softcopy display for pediatric chest radiographs are presented. The pediatric disease categories studied were pneumothorax, linear atelectasis, air bronchograms, and interstitial disease. Digital images were obtained directly from a computed radiography system. Results from the readings of 239 chest radiographs by 4 radiologists show no significant difference between viewing images on film and softcopy display for the disease categories pneumothorax and air bronchograms. A slight performance edge for softcopy was seen for the disease categories of interstitial disease and linear atelectasis.

Convolution Filtering Technique For Estimating Scatter Distributions In Radiographic Images

The degree to which beam hardening due to iodine, and x-ray and light scatter within the imaging chain, cause a nonlinear videodensitometric response of DSA grey level to iodine areal density were studied. Beam hardening effects were found to be secondary to the effects of scatter. Local scatter estimates could reliably be made, using an array of small lead beam-stops, which provided estimates of the scatter contribution regionally within an image. When the local scatter levels were subtracted from component images prior to DSA logarithmic subtraction, the DSA grey-scale response was made approximately linear with iodine areal density. Convolution filtering was studied as a method for estimating the scatter over an entire image based on beam-stop measurements made at a few points within the image. Simple, large-size (.0100x100 pixels) convolution kernels were capable of reproducing the scatter distribution within a chest image with an rms percentage error of <10% over a 36 cm field of view. The method should be implementable using a single lead beam-stop placed over a highly transmissive region in a patient.

EFFECT OF p-CHLOROMERCURIBENZENE SULFONATE ON GASTRIC PARIETAL AND SURFACE CELL FUNCTION IN THE DOG

p-Chloromercuribenzene sulfonate, a polar organic mercurial which presumably penetrates cell membranes with great difficulty, led to a profound decrease in acid recovered from exteriorized segments of dog gastric fundus stimulated by exogenous histamine. Trapping of secreted hydrogen ions by Tris buffer did not significantly reduce the back-diffusion component of apparent inhibition. Sodium-for-hydrogen exchange was close to unity before and after inhibition.