Simulated internal defibrillation in humans using an anatomically realistic three-dimensional finite element model of the thorax.

INTRODUCTION: Determination of the optimal electrode configuration during implantable cardioverter defibrillator (ICD) implantation remains largely an empirical process. This study investigated the feasibility of using a finite element model of the thorax to predict clinical defibrillation metrics for internal defibrillation in humans. Computed defibrillation metrics from simulations of three common electrode configurations with a monophasic waveform were compared to pooled metrics for similar electrode and waveform configurations reported in humans. METHODS AND RESULTS: A three-dimensional finite element model was constructed from CT cross-sections of a human thorax. Myocardial current density distributions for three electrode configurations (epicardial patches, right ventricular [RV] coil/superior vena cava [SVC] coil, RV coil/SVC coil/subcutaneous patch) and a truncated monophasic pulse with a 65% tilt were simulated. Assuming an inexcitability threshold of 25 mA/cm2 (10 V/cm) and a 75% critical mass criterion for successful defibrillation, defibrillation metrics (interelectrode impedance, defibrillation threshold current, voltage, and energy) were calculated for each electrode simulation. Values of these metrics were within 1 SD of sample-size weighted means for the corresponding metrics determined for similar electrode configurations and waveforms reported in human clinical studies. Simulated myocardial current density distributions suggest that variations in current distribution and uniformity partially explain differences in defibrillation energy requirements between electrode configurations. CONCLUSION: Anatomically realistic three-dimensional finite element modeling can closely simulate internal defibrillation in humans. This may prove useful for characterizing patient-specific factors that influence clinically relevant properties of current density distributions and defibrillation energy requirements of various ICD electrode configurations.

Fractional changes in lung capillary blood volume and oxygen saturation during the cardiac cycle in rabbits.

Changes in local pulmonary capillary blood volume (Vc) and oxygen saturation (S) have been difficult to measure in live animals. By utilizing the differences in absorption of light at two wavelengths (650 and 800 nm), we estimated the fractional change in Vc and S during the course of the cardiac cycle in eight anesthetized, ventilated rabbits at low and high lung volumes. Observations were made of the pattern of diffusely backscattered light, from an approximately 1-cm3 volume of lung illuminated with a point source placed on the pleural surface through a thoracotomy. At low lung volume, the fractional change in Vc was approximately 13%, the change in S was approximately 4.6%, and the mean S was close to 77%. The fluctuations in Vc and S lagged behind peak systemic blood pressure by about one-fifth and three-fifths of a cycle, respectively. At high lung volume, there were no important fluctuations in Vc or S, and the mean S was approximately 82%. These results are consistent with fluctuations in pulmonary capillary pressure and gas exchange over the cardiac cycle, and with decreasing capillary compliance with increasing lung volume.

Electrocardiogram trace paradoxically filtered by activation of an electrosurgical unit.

Interference in the electrocardiogram (ECG) signal in an operating room environment is common. Interference from a variety of sources, including electrosurgical units and blood warmers, have been reported. We report the occurrence of an ECG signal that was cleared of interference whenever the electrosurgical unit (ESU) was activated.

Effect of macroscopic deformation on lung microstructure.

Using an anisotropic theory of diffuse light scattering in lungs, we measured the fractional changes in geometric mean linear intercepts in orthogonal directions when freshly excised rabbit lungs were subjected to isovolume uniaxial strains. Results from the optical technique were compared with morphometric estimates of fractional changes in mean linear intercepts from the same strained and unstrained (control) lobes, with the conclusion that diffuse light scattering is adequate to estimate changes in mean free paths in different directions. We compared optical estimates of fractional changes in mean linear intercepts with the macroscopic strain field measured by displacements of pleural markers; this relationship did not significantly differ from the line of identity. We conclude that the microscopic strain field is closely matched to the macroscopic strain field during uniaxial distortion. This suggests that surface reorientation may not play a large role in the origin of the low shear modulus of the lung, but this cannot be definitively stated without comparison of these experimental results to specific model predictions of the changes in mean linear intercepts in shear deformation.

A three-dimensional finite element model of human transthoracic defibrillation: paddle placement and size.

A detailed 3-D finite element model of the conductive anatomy of the human thorax has been constructed to quantitatively assess the current density distribution produced in the heart and thorax during transthoracic defibrillation. The model is based on a series of cross-sectional CT scans and incorporates isotropic conductivities for eight tissues and an approximation of the anisotropic conductivity of skeletal muscle. Current density distributions were determined and compared for four paddle pairs and two paddle sizes. Our results show that the myocardial current density distributions resulting from a defibrillation shock were fairly uniform for the paddle pairs and sizes examined in this study. Specific details of the spatial distribution of the current density magnitudes in the heart were found to depend on paddle placement and size. When the minimum current necessary to defibrillate was delivered, the maximum myocardial current density produced with any of the paddle sizes and positions examined was less than four times the minimum current density necessary to render a myocyte in a fibrillating heart inexcitable, and less than 40% of the damage threshold. These results suggest that common clinically used defibrillation paddle positions have a safety margin as large as 2.5 for current and approximately 6 for energy.

Finite element models of thoracic conductive anatomy: sensitivity to changes in inhomogeneity and anisotropy.

A moderately detailed 3-D finite element model of the conductive anatomy of a canine thorax was used to examine the sensitivity of the results obtained during simulated transthoracic defibrillation to variations in skeletal muscle anisotropy and differing degrees of model inhomogeneity. Our results suggest that the myocardial current density distribution is not particularly sensitive to the method used to model skeletal muscle anisotropy. However, anisotropy variations caused defibrillation parameters such as paddle to paddle impedance and threshold current to change by as much as 50%. We found a greater sensitivity in the myocardial current density and the defibrillation parameters to variations in model inhomogeneity. The changes observed in both depended substantially on paddle placement. This sensitivity to paddle placement highlights the difficulty in predicting how a reduction in anatomical detail will affect the myocardial current density distribution. In general, we found the defibrillation parameters to be more sensitive than the myocardial current density distribution to the variations in anatomical detail we examined.

Geometric hysteresis in pulmonary surface-to-volume ratio during tidal breathing.

We investigated the dynamic history dependence of lung surface area-to-volume ratio (S/V) during tidal breathing in live rabbits with use of our recently developed technique of diffuse optical scattering. We also examined the effect of methacholine (continuous intravenous infusion, 1-10 micrograms.kg-1.min-1) on lung micromechanics with the same technique. Animals were anesthetized, tracheostomized, and mechanically ventilated, and the left lung was exposed through a thoracotomy. An optical fiber delivering light from a He-Ne laser was attached normal to the pleural surface, producing a circular light pattern on the pleural surface from diffusively scattered light within the parenchyma. The pattern of light intensities was measured using a CCD video camera connected to a computer. S/V during tidal breathing changed in a manner qualitatively consistent with geometric similarity. There was a small but significant hysteresis in S/V vs. volume, with S/V inspiration greater than S/V expiration at the same volume. However, during methacholine challenge, the sense of hysteresis reversed; S/V inspiration was less than S/V expiration at isovolume points. Moreover, S/V during methacholine challenge systematically decreased at all lung volumes compared with control. These findings suggest that 1) during normal tidal breathing, stress hysteresis in ductal tissue is larger than septal stress hysteresis (septal tissue plus surface tension) and 2) the effect of methacholine on tissue in the septa is greater than the corresponding effect in ductal tissue.

Effects of paddle placement and size on defibrillation current distribution: a three-dimensional finite element model.

A realistic three-dimensional finite element model of the conductive anatomy of a canine thorax was constructed for use in the study of transthoracic electrical defibrillation. The model was constructed from a series of 21 cross-sectional CT scans of a 14.5 kg beagle, each separated by 0.82 cm. The electrical conductive properties of 8 distinct tissues were incorporated, including the anisotropic properties of skeletal muscle. Current density distributions were obtained for six paddle pairings and two paddle sizes. A quantitative basis for comparing the resulting distributions was formulated. Our results suggest that placing one or both of the paddles near the heart delivers a higher fraction of current to the heart. However, such paddle placements also produce a less uniform myocardial current density distribution and thus have a higher potential for causing damage. We found that some paddle positions can produce myocardial current densities close to the threshold for damage in successful defibrillations. Results obtained for 8 and 12 cm paddles indicate that 12 cm paddles may offer modest advantages over 8 cm paddles in clinical defibrillation. Comparison of our results to available in vivo experimental data confirm the validity of the finite element method for examining continuum field variables pertinent to electrical defibrillation.

Test of four defibrillation dosing strategies using a two-dimensional finite-element model.

The most widely used defibrillation dosing strategy is that adopted by the American Heart Association in 1986. However, several alternative dosing strategies have been proposed to match delivered energy to the individual requirements of defibrillation subjects. In this study, two-dimensional finite element methods are used to investigate the performance of four of these dosing strategies applied to three thoracic models representative of men and women of different thoracic aspect ratios. From the resulting current density distributions, the relative effectiveness of the following dosing strategies are evaluated and compared: constant current; current proportional to body weight; constant energy; energy proportional to body weight. Our results show that the strategy of applying current proportional to subject body weight with a current dose of 0.58 A kg-1 was able to defibrillate all three subjects with only minimal overexposure of any one of them. None of the other dosing strategies examined could be made to successfully defibrillate all three subjects without significantly overexposing at least one.

Mechanical independence of wingbeat and breathing in starlings.

The pectoral muscles in birds comprise up to a third of the body weight and provide the principal drive to the wing. Their attachment to the sternum suggests that they could compress the thorax and assist ventilation during flight. Most, but not all, birds have an integer ratio relationship between wingbeat and breathing frequency, but no measurements of the respiratory flow associated with the act of wingbeat are available. We recorded respiratory flow and wing timing in three starlings that flew at 22 knots (11 m.s-1) for up to 5 min in a wind tunnel. Triggering on wingbeat, we ensemble averaged flow records for many wingbeats in each flight. Because wingbeats occurred throughout the respiratory cycle, breathing flow tended to average to zero, and a small flow event related to wingbeat emerged. The volume change associated with wingbeat ranged from 3 to 11% of tidal volume, and this is probably an overestimate. We conclude that wingbeat and breathing in starlings are essentially mechanically independent, despite the direct attachment of the locomotor muscles to the thorax.

Inspiratory flow rate and dead space in dogs.

It is generally accepted that a stationary concentration front is established in the tracheobronchial tree during the inspiratory phase of single- and multiple-breath washouts. The anatomic position of this front, which is determined by the balance between diffusive flux toward the airway opening and convective flux toward the periphery, is frequently used to predict the effects of molecular diffusivity and inspiratory flow rate on dead space. Although there is substantial experimental evidence supporting the predictive effect of molecular diffusivity, there is little evidence regarding the effect of convective flow. This study confirmed the predictions for the effects of molecular diffusivity but contradicted those for the effects of inspiratory flow. We measured dead space by multiple- and single-breath inert gas washout techniques and also measured physiological dead space in dogs for inspiratory flow rates of 10-71 ml.kg-1.s-1. None of the three measures of dead space increased over the entire range of flow rate, as predicted by contemporary gas transport models. A possible explanation for these findings is that axial dispersion coefficients in the anatomic region where stationary fronts are believed to develop (respiratory bronchioles and alveolar ducts) significantly increase with convective flow rate rather than remain equal to molecular diffusivity.

Diffusivity dependence of multiple-breath washouts of lung periphery.

Subpleural concentrations of He and SF6 were measured during multiple-breath washouts from isolated dog lungs. Tidal volume, inspiratory flow, and frequency were in the normal range of canine ventilation. For each gas, there was a local minimum in concentration during inspiration (Cinsp) and a local maximum in concentration during exhalation (Cexp). SF6 exhibited a deeper inspiratory trough than He for each breath of every washout. For large tidal volumes (10-20 ml/kg), Cexp approximated a single exponential decay and He was cleared more rapidly than SF6. For small tidal volumes (2.5 ml/kg), Cexp was multiexponential and SF6 was cleared more rapidly than He. Cinsp/Cexp (a measure of the depth of the inspiratory trough) and the kinetics of Cexp decay were determined for washouts using a tidal volume of 10 and 20 ml/kg and different inspiratory flows. Under all conditions, an increase of inspiratory flow resulted in a deeper inspiratory trough for both He and SF6. For washouts using 10 ml/kg and 60 breaths/min, an increase of inspiratory flow increased the clearance of both gases. In washouts using lower ventilatory frequencies, gas clearance was independent of inspiratory flow. These findings are contrary to predictions of contemporary models of convection and diffusion in the lung. This study suggests that convective axial mixing and radial diffusion in the airways are important determinants of pulmonary gas transport.

Gas conductance during high-frequency oscillatory ventilation in large animals.

Three sheep, a foal, a pony, and a calf were anesthetized and ventilated for short periods, using a high-frequency oscillatory ventilator. The efficiency of CO2 elimination was characterized at various oscillatory frequencies (50 to 30 Hz) and various tidal volumes, although the tidal volume used was always less than the measured dead space of the animal. In general, increasing either the oscillatory frequency or tidal volume increased CO2 elimination, but increasing the tidal volume had more effect. The relationship between these 3 variables was best described by a power law equation. Ventilatory frequencies and tidal volumes required to maintain eucapnia in the species studied were extrapolated from the results and, when technically possible, the potential of the technique to maintain eucapnia was tested in extended runs. The animals were supported successfully over this period, with normal blood gas tensions and no detrimental effects to heart rate and rhythm or arterial blood pressure.

Multiple gas washout during jet ventilation.

Simultaneous washouts of He, N2, and SF6 were monitored during jet ventilation with tidal volumes of 50-200 ml and rates of 1-2 Hz. Gas concentrations were measured from the trachea and from a lower lobe bronchus in six baboons by mass spectrometry. Washouts using large tidal volumes approximated single exponential decays with the relative exponential rates of decay being He fastest, SF4 slowest, and N2 intermediate. Washouts using smaller tidal volumes demonstrated a two-phase exponential decay pattern. During the fast phase, the relative exponential rates of decay were He slowest, SF6 fastest, and N2 intermediate, the reverse order seen during large-volume washouts. During the slow phase, the relative exponential rates of decay were He fastest, SF4 slowest, and N2 intermediate, the same order seen during large-volume washouts. The magnitude of the first phase observed from the lower lobe bronchus was less than that observed from the trachea. These data are consistent with a serial two-compartment transport model incorporating a limitation of molecular diffusion between the peripheral and proximal compartments. The more rapid clearance of less diffusible gases from the central airways during the first phase of washout was due to slower transport from the alveoli to the central airways rather than faster transport from the central airways to the airway opening.

Do's and don'ts of installing a RIS.

This report reviews several important issues relating to the acquisition of a radiology information system (RIS). It emphasizes the importance of defining specific goals for computerizing a department and for understanding that department's manual operations. It then outlines methods to request proposals from prospective vendors and to evaluate their responses. Suggestions for negotiating a favorable arrangement with a vendor follow. Finally, the summary lists five do's and three don'ts for selecting and installing a RIS.

Safe and cost-effective cleaning of pressure-monitoring transducers.

The CDC category I recommendation for high-level disinfection of pressure-monitoring transducers between uses imposes substantial costs. This practice has not been shown to be useful when disposable transducer domes are used. A prospective, randomized, double-blind study assessed the consequences of wiping transducers with alcohol between uses rather than sterilizing with ethylene oxide. We studied 5,197 transducer courses with disposable domes involving 2,202 patients in intensive care units during 15 months. There was no significant difference between the two treatment regimens in the risk of positive cultures of fluids from monitored lines, of cultures of cannula tips or of a positive blood culture. We conclude that under normal conditions, ie, in the absence of a cluster of transducer-related bacteremias, wiping transducers with alcohol between uses is sufficient when disposable domes are used.

ACR-NEMA digital imaging and communications standards: minimum requirements.

The purposes, implications, and history of development of the American College of Radiology-National Electrical Manufacturers Association (ACR-NEMA) Digital Imaging and Communication Standard and its contents are briefly described, and the minimum requirements of the ACR-NEMA Digital Imaging and Communication Standard are described with a concise introduction of each layer. The usefulness, validity, current status, and future development of the standard are also discussed.

Inspiratory aerodynamic valving in goose lungs depends on gas density and velocity.

The non-reversing gas flow pattern in the avian lung has been attributed to 'aerodynamic valves'. A fundamental property of all aerodynamic valves is their dependence on inertial forces in the gas stream: sufficient reduction of inertial forces will cause aerodynamic valves to fail. If valving in the avian lung is aerodynamic, it should fail when gas stream momentum is reduced. We tested the dependence of the inspiratory valves in the goose lung on gas density and gas flow velocity. A bolus of tracer gas was placed in the tracheal cannula during an end-expiratory pause. Tracer gas appearance in a cranial air sac during the following inspiration and pause was used to deduce failure of the 'inspiratory valve' in cyclically ventilated geese. Little or no tracer entered the sac under control conditions, which approximated resting breathing, indicating highly effective valving. Lower flow rate or lower gas density caused increased tracer appearance, indicating valve failure. These results demonstrate the importance of gas inertial forces to normal valve function, and are direct evidence for the aerodynamic nature of the avian inspiratory valve.

Longitudinal elastic wave propagation in pulmonary parenchyma.

Consideration of the lung as an elastic continuum led us to investigate the possible propagation of elastic waves. Here the relevant stiffness and density are given by the Lamé constants and density of the parenchyma. To test this hypothesis, we measured propagation velocities (c) in dog lobes by recording transit times of a velocity impulse on one side of the lobe and the subsequent arrival on the other side. We compared our measured values of c with elastic longitudinal wave velocities (c long) predicted by values of elastic moduli given by Lai-Fook et al. (J. Appl. Physiol. 40: 508-513, 1976) as a function of translobar pressure (PL) and our measured densities. Good agreement was found between c and c long. Typical values of c ranged from 250-1,500 cm/s as PL ranged from 2-20 cmH2O. No systematic difference in the c-c long relation was found between inflation and deflation, suggesting that the elastic moduli of lungs are essentially a function of pressure. No significant effect was observed by changing the physical properties of the gas within the lobe [air vs. He vs. sulfur hexafluoride (SF6)], suggesting that indeed we were observing waves associated with the coupling of parenchymal density to parenchymal stiffness.