Effects of hypoxia on pulmonary microvascular volume.

To determine the effects of alveolar hypoxia on pulmonary microvascular volume, X-ray microfocal angiographic images of isolated perfused dog lung lobes were obtained during passage of a bolus of radiopaque contrast medium during both normoxic (alveolar gas, 15% O(2), 6% CO(2), and 79% N(2)) and hypoxic (3% O(2), 6% CO(2), and 91% N(2)) conditions. Regions of interest (ROIs) over the lobar artery and vein at low magnification and a feeding artery ( approximately 500 microm diameter) and the nearby microvasculature (vessels smaller than approximately 50 microm) at high magnification were identified, and X-ray absorbance vs. time curves were acquired under both conditions from the same ROIs. The total pulmonary vascular volume was calculated from the flow and the mean transit time for the contrast medium passage from the lobar artery to lobar vein. The fractional changes in microvascular volume were determined from the areas under the high-magnification X-ray absorbance curves. Hypoxia decreased lobar volume by 13 +/- 3% (SE) and regional microvascular volume by 26 +/- 4% (SE). Given the morphometry of the lung vasculature, these results suggest that capillary volume was decreased by hypoxia.

Transit time dispersion in the pulmonary arterial tree.

Knowledge of the contributions of arterial and venous transit time dispersion to the pulmonary vascular transit time distribution is important for understanding lung function and for interpreting various kinds of data containing information about pulmonary function. Thus, to determine the dispersion of blood transit times occurring within the pulmonary arterial and venous trees, images of a bolus of contrast medium passing through the vasculature of pump-perfused dog lung lobes were acquired by using an X-ray microfocal angiography system. Time-absorbance curves from the lobar artery and vein and from selected locations within the intrapulmonary arterial tree were measured from the images. Overall dispersion within the lung lobe was determined from the difference in the first and second moments (mean transit time and variance, respectively) of the inlet arterial and outlet venous time-absorbance curves. Moments at selected locations within the arterial tree were also calculated and compared with those of the lobar artery curve. Transit times for the arterial pathways upstream from the smallest measured arteries (200-micron diameter) were less than approximately 20% of the total lung lobe mean transit time. Transit time variance among these arterial pathways (interpathway dispersion) was less than approximately 5% of the total variance imparted on the bolus as it passed through the lung lobe. On average, the dispersion that occurred along a given pathway (intrapathway dispersion) was negligible. Similar results were obtained for the venous tree. Taken together, the results suggest that most of the variation in transit time in the intrapulmonary vasculature occurs within the pulmonary capillary bed rather than in conducting arteries or veins.

Regional perfusion parameters from pulmonary microfocal angiograms.

An indicator-dilution model was developed to describe transport of vascular contrast medium through an organ during acquisition of vascular dynamic contrast images. The model provides the theoretical basis for methods of determining regional blood flow, blood volume, and mean transit time from time-absorbance curves acquired from the images of tissue regions of interest (ROI) distal from the inlet site. The robustness of these methods was evaluated using a computer-simulated vessel network, which simulated the passage of a bolus of contrast medium through arterioles, networks of capillaries, and venules. The network was used to evaluate the reliability of ROI parameter estimation methods when the underlying model assumptions are violated. The shape of the ROI inlet concentration curve and moderate amounts of random noise did not affect the ability of the method to recover accurate parameter estimates. The estimates of ROI flow and transit time were degraded in the presence of significant dispersion of the inlet concentration curve as it traveled through arteries upstream from the microvascular ROI or when the flow was redistributed within the ROI. The estimates of ROI volume were relatively robust. The method was applied to image data of the dog pulmonary vasculature obtained using microfocal X-ray angiography to show that the results obtained from the simulations are consistent with actual data.

Edema development and recovery in neurogenic pulmonary edema.

We determined the time course of changes in extravascular lung water (EVLW) that occur after massive sympathetic activation produced by intracisternal veratrine administration in chloralose-anesthetized dogs. Three groups of dogs were studied. In the first group (n = 9), acute increases in EVLW (occurring within minutes) were determined both by measuring extravascular thermal volume and by gravimetric analysis. In the second (n = 6) and third (n = 7) groups, changes in EVLW were followed for 2-3 h after veratrine administration. Extravascular thermal volume was measured in the second group. In the third group, right atrial injections of a vascular indicator (125I-labeled serum albumin) and an extravascular indicator (3HOH) were made while blood was sampled from the pulmonary artery (PA) and left atrium, and EVLW was determined by deconvolution of the left atrial and PA concentration-time curves. Indicator-dilution and gravimetric EVLW increased acutely only in dogs in which PA pressure exceeded 60 Torr, with two- to four-fold increases in EVLW being observed in dogs that developed the highest PA pressures (maximum 94 Torr). Thus, severe edema can develop rapidly after massive sympathetic nervous system activation but requires extreme degrees of pulmonary hypertension. In several dogs after the acute increase in EVLW associated with the pulmonary hypertension, the indicator-dilution EVLW decreased with time. These decreases appear to effect clearance of edema fluid rather than alterations in perfusion.

Regional transit time estimation from image residue curves.

Methods for estimating regional flow from digital angiography or dynamic computed tomography images require determination of indicator mean transit time (t) through a region-of-interest (ROI). We examine how the ROI kinematics and input dispersion influence the recovery of t using a computer-simulated vessel network representing that which might occur in a real organ. The network simulates flow through a large artery branching into two small arteries, each feeding a system of smaller vessels intended to represent capillaries and small vessels below the resolution of the imaging system. The capillaries are drained by a similar system of veins. Concentration curves measured over the inlet to the network and microvascular ROI residue curves are simulated. When the area-height ratio of the microvascular ROI curve is used and all of the indicator is contained within the ROI for at least one time point, t is recovered exactly. As the size of the ROI is reduced or the inlet concentration curve becomes more dispersed, the error in the recovery of t grows. By first deconvolving the inlet concentration curve from the microvascular ROI curve, and then calculating the area-height ratio, t is recovered accurately. If the inlet concentration curve becomes more dispersed between its measured site and the actual inlet to the ROI, or if the flow distribution within the ROI is changed, the estimation of t can be degraded. To put the simulations in perspective relative to an example of image data, the methods were applied to microfocal x-ray angiography data obtained from a approximately 700 micron canine pulmonary artery and vein, the surrounding microvasculature and the inlet lobar arterial cannula.

Model-free numerical deconvolution of recirculating indicator concentration curves.

This paper investigates two model-free methods for numerical deconvolution of recirculating indicator concentration curves. The two methods, damped least squares and discrete orthogonal polynomial deconvolution, are applied to simulated data to verify the reliability of the algorithms. Both deconvolution methods provide damping that results in estimated transport functions that are smooth and reasonable estimates of the actual simulated transport function. On convolution with the simulated input curve, the estimated transport functions provide good fits to the simulated output curve. In addition, methods for identifying an optimal solution and for truncating the artifactually long oscillatory tails of the estimated transport functions are proposed, which appear to allow for reasonably accurate estimation of the mean transit times and variances of the transport functions as well. When either method was applied to indicator dilution data obtained from the pulmonary artery and left atrium, it was computationally stable while producing transport functions that when convolved with the input concentration curves provided good fits to the output concentration curves. The combined simulation and experimental results suggest that the proposed methods should be useful for estimating circulation transport functions from indicator dilution data.

Distensibility of small veins of the dog lung.

To determine the distensibility of the intrapulmonary veins (250-2,900 microns diam) of the dog lung, we obtained X-ray angiograms from isolated lung lobes over a vascular pressure range of approximately 0-30 Torr. Over this pressure range the diameter vs. pressure curves tended to flatten out at the high pressures. In the pressure range of 0-19 Torr, we characterized the vessel distensibility by alpha (the ratio of the slope, beta, of the graph of diameter vs. intravascular pressure to the intercept, Do). The average value of alpha was approximately 1.2%/Torr. There was a weak negative correlation (r = -0.32) between alpha and Do. Infusion of enough norepinephrine to produce approximately 50% increase in total lobar vascular resistance produced a decrease in Do and alpha of approximately 33 and 32%, respectively.

An angiographic method for in vivo study of arteries of the circle of Willis in small animals.

An X-ray imaging technique designed to allow sequential diameter measurements of the cerebral vessels in intact, anesthetized small animals under relatively physiological conditions is described. The ferret and the rabbit were chosen as potentially useful animal models for studying the cerebrovascular system because of the advantageous anatomic characteristics of these relatively small species. A commercially available and relatively inexpensive X-ray imaging system with a small focal spot provides good spatial resolution. An external carotid perfusion loop allows for 1) the introduction of low-osmolality contrast medium without changing perfusion pressure or flow and 2) measurement of internal carotid and circle of Willis pressures at the same time that the vessel images are obtained. In the present study, detection of small changes in the diameters of the small vessels is facilitated by an algorithm utilizing the X-ray absorption by the entire vessel cross section. This avoids some of the problems of edge detection for small cylindrical vessels wherein the contrast is less than optimal and diminishes as the vessel perimeter is approached.

An algorithm for angiographic estimation of blood vessel diameter.

This study was carried out in an attempt to develop an objective and robust method for measuring changes in the diameters of small blood vessels from X-ray angiographic images. Recognizing potential problems with edge detection methods applied to cylindrical vessels in which the contrast diminishes as the boundary is approached, we have attempted to utilize the X-ray absorbance data across the entire cross section of the vessel. Then, assuming a cylindrical geometry, the absorbance data are fit to the cylindrical absorbance function by use of nonlinear regression analysis. The method was tested and calibrated using glass tubes filled with various concentrations of contrast medium. The diameters of small pulmonary arteries were estimated by applying the method of angiograms obtained from an isolated dog lung lobe. The structure of the residuals obtained after the fitting procedure was analyzed to test the appropriateness of the model for use with images of vessels. The results suggest that this approach will have utility for systematically quantifying vessel dimensions.