Hypoxia

Background: Electrical impedance tomography (EIT) is a noninvasive imaging technique using impedance to visualize and measure blood volume changes.

Study objective: To examine the validity of EIT in the measurement of hypoxic pulmonary vasoconstriction (HPV) and hyperoxic pulmonary vasodilation in healthy volunteers and COPD patients.

Participants: Group 1 consisted of seven healthy volunteers (mean age, 46 years; age range, 36 to 53 years). Group 2 comprised six clinically stable COPD patients (mean age, 65 years; age range, 50 to 74 years).

Interventions: EIT measurements were performed in healthy subjects while they were breathing room air, 14% oxygen (ie, hypoxia), and 100% oxygen (ie, hyperoxia) through a mouthpiece. Maximal impedance change during systole ([DELTA]Zsys) was used as a measure of pulmonary perfusion-related impedance changes. Stroke volume (SV) was measured by means of MRI. In the COPD group, EIT and SV also were determined, but only in room air and under hyperoxic conditions.

Results: The data were statistically compared to data for the room air baseline condition. In the volunteers, the mean ([+ or -] SD) [DELTA]Zsys for the group was 352 [+ or -] 53 arbitrary units (AU) while breathing room air, 309 [+ or -] 75 AU in hypoxia (p < 0.05), and 341 [+ or -] 69 AU in hyperoxia (not significant [NS]). The mean MRI-measured SV was 83 [+ or -] 21 mL while breathing room air, 90 [+ or -] 29) mL in hypoxia (NS), and 94 [+ or -] 19 mL in hyperoxia (p < 0.05). In the COPD patients, the mean [DELTA]Zsys for this group was 222 [+ or -] 84 AU while breathing room air and 255 [+ or -] 83 AU in hyperoxia (p < 0.05). In this group, the SV was 59 [+ or -] 16 mL while breathing room air and 61 [+ or -] 13 mL in hyperoxia (NS). Thus, the volunteer EIT response to hypoxia is not caused by decreased SV, because SV did not show a significant decrease. Similarly, in COPD patients the EIT response to hyperoxia is not caused by increased SV, because SV showed only a minor change.

Conclusion: EIT can detect blood volume changes due to HPV noninvasively in healthy subjects and hyperoxic vasodilation in COPD patients.

Key words: electrical impedance tomography; noninvasive; COPD; hypoxia; hyperoxia; pulmonary vasoconstriction; dilation

Abbreviations: AU = arbitrary units; DLCO = diffusing capacity of the lung for carbon monoxide; EIT = electrical impedance tomography; HPV = hypoxic pulmonary vasoconstriction; NS = not significant; ROI = region of interest; Sp[O.sub.2] = pulse oximetric saturation; SV =stroke volume; [DELTA]Zsys = maximal impedance change during systole as a measure of pulmonary perfusion-related impedance changes

**********

Constriction or dilation of the pulmonary vascular bed makes it possible to respond and adapt to many different physiologic conditions. (1) In many clinical situations, the pulmonary vascular bed is damaged, and this condition may interfere with this adaptation. For instance, in COPD patients the size of the pulmonary vascular bed is significantly reduced, and structural changes of the microvascular vessels, due to chronic hypoxic pulmonary vasoconstriction (HPV), occur. (2-8)

Long-term oxygen administration has been proposed as a method with which to prevent these structural changes and, by doing so, to prevent the development of pulmonary artery hypertension. (9,10) However, not all COPD patients respond to oxygen therapy. (11,12) It has been hypothesized that those patients with reversible chronic HPV will benefit more from the continuous administration of oxygen. (12) Therefore, it would be useful to find a technique that makes it possible to detect an oxygen-induced release of HPV.

Electrical impedance tomography (EIT) is a technique that makes it possible to measure dynamic changes of the amount of blood in the pulmonary vascular bed during the cardiac cycle. (13-16) The principle of EIT in the measurement of pulmonary perfusion is based on the visualization of impedance changes inside the thorax due to changes in the blood volume during systole, relative to the blood volume during diastole within the pulmonary vascular bed. HPV reduces the ability of the blood vessels to distend, and consequently hypoxia causes a reduction of the impedance change during systole as measured by EIT. Although EIT is able to detect HPV in healthy subjects, it remains unknown whether EIT can detect relaxation of HPV in COPD patients. (16) Therefore, the aim of this study was to examine the validity of EIT in the measurement of HPV and the relaxation of HPV in healthy volunteers and COPD patients, respectively.

MATERIALS AND METHODS

Patients

The study comprised the following two groups: group 1 consisted of seven healthy volunteers (six men and one woman; mean age, 46 years; age range, 36 to 53 years) without any sign of cardiac or respiratory diseases and with normal lung function. Group 2, the COPD patient group (five men and one woman; mean age, 65 years; age range, 50 to 74 years), consisted of clinically stable emphysema patients. The diagnosis of COPD was made according to the criteria of the American Thoracic Society. (17) The mean FE[V.sub.1] in the patient group was 1,920 mL (FE[V.sub.1] range, 1,010 to 2,190 mL or 37 to 88% of predicted; mean FE[V.sub.1] 66% of predicted). The mean FE[V.sub.1]/vital capacity ratio was 0.44 (range, 0.30 to 0.65). Furthermore, the mean diffusing capacity of the lung for carbon monoxide (DLCO) in the patient group was 53% of predicted (range, 37 to 64% of predicted).

The research protocol was approved by the institutional human ethics committee. All patients and volunteers gave informed consent.

Study Design

The first part of the study was designed to evaluate the influence of hypoxia on the maximal systolic impedance change in the lung, as a consequence of pulmonary vasoconstriction due to hypoxia in healthy subjects. Since changes in stroke volume (SV) during hypoxia might influence the EIT outcomes, MRI measurements were performed to measure changes in SV due to hypoxia. For technical reasons, MRI measurements and EIT measurements could not be performed simultaneously. First, the healthy subject was positioned in the MRI apparatus. After a stabilizing period of 10 rain, MRI measurements were made. The healthy subject then started inhaling 14% oxygen. This gas mixture was provided via a Douglas bag that was connected to the mouthpiece. The subject had a clip on his nose to ensure that he was breathing only the hypoxic oxygen mixture. After 10 rain of stabilization, the second SV measurement was performed. Pulse oximetric saturation (Sp[O.sub.2]) was measured continuously by pulse oximetry on a finger, and from the saturation values the Pa[O.sub.2] was calculated for the various breathing conditions. Two days after obtaining the MRI measurements, the whole procedure was repeated with EIT. All measurements, with EIT as well as Mill, were performed under equal conditions. Subjects were in the supine position with their arms stretched above their head to enlarge the distance between EIT-electrodes and the heart. Measurements were started after acclimatization and stabilization of heart rate, respiratory rate, and Sp[O.sub.2]. Subjects were not allowed to eat, drink, or smoke before and during the protocol.

In the second part of the study, we examined the relaxation of HPV in healthy subjects and COPD patients. After a wash-in period of 10 min breathing 100% oxygen, MRI measurements were performed. EIT measurements were repeated in the same way 2 days later.

EIT Protocol

The EIT measurements were performed with an applied potential tomograph device (model DAS-01 P Portable Data Acquisition System; IBEES; Sheffield, UK). (18,19) Measurements were performed with 16 Ag/AgCl electrodes (Meditrace 200; Graphic Controls; Gananoque, ON, Canada) equidistantly attached in a transverse plane at the level of the third intercostal space. During the EIT measurements, the subjects lay in the supine position with their arms above their head. A source of current that generated a monofrequent sinusoidal current (ie, 50 kHz and 5 mA peak to peak) was used to inject the current into the body. Impedance measurements were processed by means of the Sheffield back-projection algorithm to yield images. Data collection was synchronized with the R wave of the ECG. Two hundred cardiac cycles were averaged to obtain one complete data set. Breathing artifacts were automatically removed by the averaging procedure. One data set contained 30 frames, spaced 40 ms apart.

As difference images were generated with the applied potential tomograph device, the first frame was defined as the reference data set in the present study. Since the resistivity of blood is less than that of other tissues, the presence of blood results in a decrease of the measured impedance. (20) This makes it possible to study the dynamics of the pulmonary blood volume changes during the cardiac cycle in the sequence of images produced by the ECG-gated EIT. In this study, the maximal impedance change during systole was measured as pulmonary perfusion-related impedance changes ([DELTA]Zsys). All measurements of impedance change should be expressed as the value x [10.sup.-5], as can be seen in the figures. For clarity, this factor is not mentioned in the results.

To quantify the impedance change within the lungs, region-of-interest (ROI) analysis was performed. A specific area (ie, the inner half circle) was chosen as the ROI to exclude impedance increase in the anterior zone of the thorax (probably due to the heart) and to exclude disturbance at the borders of the image (Fig 1). This ROI was the same in all measurements. The average pixel value within the ROI was plotted as a function of time to show the impedance change during the cardiac cycle. The average pixel value has no unit since it is dimensionless as a consequence of the reconstruction algorithm based on normalized differences. Therefore, the change in the average pixel value in the sequence during the cardiac cycle relative to end-diastole was expressed as an arbitrary unit (AU).

[FIGURE 1 OMITTED]

MRI Velocity Quantification in the Pulmonary. Artery

MRI was performed on a 1-T whole-body system (Impact Expert; Siemens; Erlangen, Germany), using a quadruple, phased-array, circularly polarized body coil, with two receiver antennas locally applied on the anterior chest wall, and two on the posterior chest wall.

MRI velocity mapping was performed on the main pulmonary artery. A single oblique image plane was planned on a sagittal scout image that showed the right ventricle and the main pulmonary artery arising from it. A two-dimensional gradient-echo pulse sequence was used with excitation angle of 25[degrees], an echo time of 6.5 ms, and a receiver bandwidth of 195 Hz. One-dimensional velocity encoding was perpendicular to the image plane. The phase-encoding steps of two different acquisitions (repetition time, 14 ms) were interleaved, one with velocity encoding of the phase, and one without. Subtraction of the resulting phase maps compensated for phase changes caused by inhomogeneity of the magnetic field, leaving only phase changes related to velocity. The temporal resolution within the cardiac cycle was thus 28 ms (ie, 2 x 14 ms). The velocity sensitivity was set at 150 cm/s by proper adjustment of the amplitude of the velocity-encoding gradients. The field of view was 300 [mm.sup.2], and the matrix size was 230 x 256. The RR interval (ie, the time between two heartbeats) was automatically registered during MRI acquisition of the main pulmonary artery flow.

Analysis of Flow Curves

The main pulmonary artery flow curve was evaluated as follows. In each time phase of the velocity images, the cross-sectional area of the artery was delineated by hand in order to account for translations of the artery with respect to the image plane. The spatial mean velocity in this area was plotted against time. No aliasing due to high peak systolic velocities was encountered. Volume flow was obtained by multiplying the spatial mean velocity with the cross-sectional area. Finally, integrating the volume-flow curve over systole yielded right ventricular SV. (21)

Statistical Analysis

The Mann-Whitney test was used for comparing the healthy subject group and the COPD group. The Wilcoxon signed rank test for matched pairs was used to compare [DELTA]Zsys during hypoxia and while breathing room air, and to compare [DELTA]Zsys during hyperoxia and while breathing room air. Analyses were performed with a statistical software package (GraphPad Prism, version 3.02 for Windows; GraphPad Software; San Diego, CA). All results were reported as the mean [+ or -] SD. A p value of < 0.05 was considered to be significant.

RESULTS

Adequate EIT and MRI measurements could be performed in every subject. The mean impedance in the healthy subjects was 352 [+ or -] 53 AU, and in the COPD group it was 222 [+ or -] 84 AU (p < 0.05). There was a significant correlation (r = 0.73; p = 0.005) between DLCO and [DELTA]Zsys under room air conditions (Fig 2). The individual Sp[O.sub.2] values for the healthy subjects are represented in Table 1, and for the COPD patients in Table 2. For the room air condition and hypoxia, the corresponding Pa[O.sub.2] values, calculated from the Sp[O.sub.2], also are provided.

[FIGURE 2 OMITTED]

Figure 3, top left, A, shows the ECG-gated impedance change in seven healthy subjects when they were breathing room air and when they were breathing a hypoxic mixture with 14% oxygen. In six subjects, a decrease in impedance change was noted, and in one subject a slight increase in impedance change was noted, when they were breathing 14% oxygen. The mean impedance decrease was significant (309 [+ or -] 75 vs 352 [+ or -] 53 AU; p < 0.05). During hypoxia, the mean Pa[O.sub.2] decreased from 101 [+ or -] 12 to 59 [+ or -] 4 mm Hg (p < 0.05). Figure 3, top right, B, and bottom left, C, show the SVs and heart rates respectively, as measured by MRI, of the same seven subjects when they were breathing room air or 14% oxygen. Five subjects showed an increase of SV during hypoxia, and in two subjects there was a decrease. The SV was 83 [+ or -] 21 mL while breathing room air, and 90 [+ or -] 29 mL while breathing 14% oxygen (not significant [NS]). The mean heart rate in this healthy subject group was 72 [+ or -] 19 beats/min while breathing room air, and increased to 88 [+ or -] 18 beats/min while breathing the hypoxic gas mixture (p < 0.05).

[FIGURE 3 OMITTED]

The systolic impedance change when healthy subjects and COPD patients breathed room air or 100% oxygen are shown in Figure 4, top left, A. When the healthy subjects breathed 100% oxygen, the mean impedance change (341 [+ or -] 69 AU) was not significantly different from that while breathing room air (352 [+ or -] 53 AU). On the contrary, when six emphysema patients were breathing 100% oxygen, the mean impedance change was 255 [+ or -] 83 AU, while this was 222 [+ or -] 84 AU when they were breathing room air (p < 0.05). The mean Sp[O.sub.2] in the healthy subject group was 97% while breathing room air, while the Sp[O.sub.2] was 99% when they were breathing 100% oxygen. In the COPD group, these values were 96% and 99%, respectively (p < 0.05). SVs and heart rates in healthy subject and patient groups inhaling room air and 100% oxygen are represented in Figure 4, top right, B, and bottom left, C, respectively. In the healthy group, SV increased from 83 [+ or -] 21 to 94 [+ or -] 19 mL (p < 0.05), while in the COPD patient group SV was 59 [+ or -] 16 mL when breathing room air and 61 [+ or -] 13 mL when breathing 100% oxygen (NS). Heart rate increased from 72 [+ or -] 19 to 84 [+ or -] 16 beats/min in the healthy group when 100% oxygen was inhaled (p < 0.05), and remained unchanged in the COPD group (76 [+ or -] 13 vs 76 [+ or -] 8 beats/rain; p = NS).

[FIGURE 4 OMITTED]

DISCUSSION

EIT can be used as a technique that uses changes in impedance to visualize and measure blood volume changes in the lung. (16) Until now, the influence of only hypoxia, and not of hyperoxia, has been investigated, and the possible influence of change in SV on the EIT signal had not yet been studied. Although a comparison between EIT measurements and invasive measurements has been made, showing a correlation between pulmonary vascular resistance and [DELTA]Zsys, these invasive measurements do not measure dynamic blood volume changes in the pulmonary vessels and, thus, should not be considered as a "gold standard" by which to validate the EIT method. (22) Therefore, we used DLCO measurements as an index of the pulmonary vascular bed and well-known physiologic responses on hypoxia and hyperoxia to validate this technique for the measurement of blood volume changes in the lung.

We were able to show a significant correlation between DLCO and [DELTA]Zsys. However, one should realize that blood volume changes in the pulmonary vascular bed (ie, what is measured by means of EIT) contribute only a small part to DLCO.

The physiologic responses during hypoxia and hyperoxia were measured in healthy subjects and COPD patients. Since changes in SV might influence the interpretation of the results, we measured SV changes by means of MRI. For ethical reasons, the hypoxic mixture was given only to healthy subjects.

The results show that there was a significant decrease in impedance during hypoxia in the healthy subjects. This can be explained as follows. The blood volume changes measured by means of EIT are mainly caused by the distensibility of the small blood vessels due to the systolic pulse wave, since these vessels contain the largest blood volume in the lungs. Since HPV in the acute phase is the only known mechanism that controls the distensibility of these arterioles, HPV will influence the EIT signal under physiologic circumstances. It might be hypothesized that SV changes also might influence the signal. An increase in SV then will be reflected as an increase in [DELTA]Zsys. However, SV did not change during hypoxia, whereas [DELTA]Zsys decreased. Therefore, our results could not be explained by changes in SV. The lack of correlation between SV and systolic change in pulmonary blood volume (as measured by EIT) can be explained as follows: systolic pulmonary blood flow will lead to a distension of the pulmonary vascular bed and an increase in blood velocity. Only this distension of the pulmonary vessels will cause a volume change of the blood vessels and thus a change in electrical impedance. (16) Thus, a decrease in SV will not change the EIT signal if this is not accompanied by a change in the distensibility characteristics of the pulmonary microvascular bed. There was a significant increase in heart rate between breathing room air and breathing 14% oxygen (ie, hypoxia). But impedance change is independent from heart rate as it is measured per heart cycle, comparing maximal impedance at systole relative to end-diastole. Furthermore, it has been shown that there is a large interindividual variability in EIT changes due to hypoxia. This might be explained by the fact that there is a large interindividual difference in sensitivity to hypoxic stimuli. (23,24)

The second part of the study examined the influence of hyperoxia in healthy subjects and in COPD patients. Sp[O.sub.2] increased when healthy subjects inhaled 100% oxygen instead of room air, as might be expected. However, this will not provide an impedance increase, because healthy subjects will not have HPV under room air condition. As the Pa[O.sub.2] was calculated from the measured Sp[O.sub.2], during hyperoxia one measures on the horizontal part of the oxygen-saturation curve, and at these high oxygen saturation values the real Pa[O.sub.2] cannot be calculated correctly. For this reason, only the measured Sp[O.sub.2] values during hyperoxia and not the Pa[O.sub.2] values are provided in Tables 1 and 2. We measured the arterial oxygen saturation noninvasively instead of performing blood gas analysis, as it was not our purpose to measure, by EIT, at which Pa[O.sub.2] value HPV started or disappeared. The goal of the study was to investigate whether HPV became apparent, and could be detected by EIT, during the inhalation of a hypoxic gas mixture by healthy individuals, or whether HPV was relieved during the inhalation of a hyperoxic gas mixture in COPD patients.

We found an increase in heart rate while patients breathed 100% oxygen, where others found a decrease during hyperoxia. (25,26) Since we could not find a physiologic explanation for this increase in heart rate, we think that discomfort of the mouthpiece, secretion of saliva while in the supine position, and a dry throat due to the pure [O.sub.2] might be the cause.

Since only COPD patients might have hypoxic vasoconstriction under room air conditions, which is inherent in their disease, 100% oxygen will cause vasodilation only in these patients and not in healthy subjects. Our results indeed showed an impedance increase in the COPD patients and not in the healthy subjects, indicating that there was only a vasodilative response to breathing 100% oxygen in the COPD group. Again, our results could not be explained by changes in SVs, as these remained unchanged.

This indicates that EIT is a sensitive method with which to detect HPV and the relaxation of HPV. This might be a very valuable instrument in clinics, as some (but not all) COPD patients react to oxygen with vasodilation. Theoretically, only the patients who have chronic HPV and are still in a reversible stage will benefit from long-term oxygen supplementation. Ashutosh and coworkers (12) conducted a study with 28 COPD patients and provided them with 28% oxygen for 24 h. They were able to divide those patients into a responding group and a nonresponding group, in which response was defined as a minimal fall in the mean pulmonary artery pressure of 5 mm Hg. After catheterization, all subjects were prescribed supplemental oxygen at a rate of 2 L/min by nasal cannula. The authors reported a strong, significant, 2-year survival benefit and improvement of quality of life in the responders group. Moreover, there was no improvement in mortality in the nonresponding group in comparison with patients who had not been treated with long-term domiciliary oxygen therapy. (12) So, in terms of the release of HPV with supplemental oxygen, it is important to select those COPD patients who will benefit the most from long-term oxygen therapy, and to prevent others from inconvenience and unnecessary expenses. EIT might be a suitable technique for selecting those patients in a noninvasive way.

In conclusion, this study showed that EIT is able to detect blood volume changes related to HPV in healthy subjects and vasodilation due to the relaxation of HPV in COPD patients. The clinical consequences and the prognostic value of these findings should be clarified in additional studies.

REFERENCES

(1) Voelkel NF. Mechanisms of hypoxic pulmonary vasoconstriction. Am Rev Respir Dis 1986; 133:1186-1195

(2) Shelton DM, Keal E, Reid L. The pulmonary circulation in chronic bronchitis and emphysema. Chest 1977; 71(suppl): 303-306

(3) Reid LM. Structure and function in pulmonary hypertension: new perceptions. Chest 1986; 89:279-288

(4) Davies P, Maddalo F, Reid L. Effects of chronic hypoxia on structure and reactivity of rat lung microvessels. J Appl Physiol 1985; 58:795-801

(5) Wright JL, Petty T, Thurlbeck WM. Analysis of the structure of the muscular pulmonary arteries in patients with pulmonary hypertension and COPD: National Institutes of Health nocturnal oxygen therapy trial. Lung 1992; 170:109-124

(6) Hillier SC, Graham JA, Hanger CC, et al. Hypoxic vasoconstriction in pulmonary arterioles and venules. J Appl Physiol 1997; 82:1084-1090

(7) Dumas JP, Bardou M, Goirand F, et al. Hypoxic pulmonary vasoconstriction. Gen Pharmacol 1999; 33:289-297

(8) Fishman AP. Chronic cor pulmonale: state of the art. Am Rev Respir Dis 1976; 114:775-794

(9) Oswald-Mammoser M, Weitzenblum E, Quoix E, et al. Prognostic factors in COPD patients receiving long-term oxygen therapy. Chest 1995; 107:1193-1198

(10) Weitzenblum E, Oswald M, Apprill M, et al. Evolution of physiological variables in patients with chronic obstructive pulmonary disease before and during long-term oxygen therapy. Respiration 1991; 58:126-131

(11) Weitzenblum E, Oswald M, Mirhom R, et al. Evolution of pulmonary haemodynamics in COLD patients under long-term oxygen therapy. Eur Respir J 1989; 2:669s-673s

(12) Ashutosh K, Mead G, Dunsky M. Early effects of oxygen administration and prognosis in chronic obstructive pulmonary disease and cor pulmonale. Am Rev Respir Dis 1983; 127:399-404

(13) McArdle FJ, Suggett AJ, Brown BH, et al. An assessment of dynamic images by applied potential tomography for monitoring pulmonary perfusion. Clin Phys Physiol Meas 1988; 9:87-91

(14) Zadehkoochak M, Blott BH, Hames TK, et al. Pulmonary perfusion and ventricular ejection imaging by frequency domain filtering of EIT images. Clin Phys Physiol Meas 1992; 13:191-196

(15) Brown BH, Barber DC, Morice AH, et al. Cardiac and respiratory related electrical impedance changes in the human thorax. IEEE Trans Biomed Eng 1994; 41:729-734

(16) Vonk Noordegraaf A, Kunst PWA, Janse A, et al. Pulmonary perfusion measured by means of electrical impedance tomography. Physiol Meas 1998; 19:263-273

(17) American Thoracic Society. Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1995; 152:S77-S120

(18) Smith RWM, Freeston IL, Brown BH. A real time electrical impedance tomography system for clinical use: design and preliminary results. IEEE Trans Biomed Eng 1995; 31:133-240

(19) Brown BH, Seagar AD. The Sheffield data collection system. Clin Phys Physiol Meas 1987; 8A:91-97

(20) Duck FA. Physical properties of tissue: a comprehensive reference book. London, UK: Academic Press, 1990

(21) Marcus JT, Vonk-Noordegraaf A, De Vries PMJM, et al. MRI evaluation of fight ventricular pressure overload in chronic obstructive pulmonary disease. J Magn Reson Imaging 1998; 8:999-1005

(22) Smit HJ, Vonk-Noordegraaf A, Roeleveld RJ, et al. Epoprostenol-induced pulmonary vasodilatation in patients with pulmonary hypertension measured by electrical impedance tomography. Physiol Meas 2002; 23:237-243

(23) Weitzenblum E, Sautegeau A, Ehrhart M, et al. Long-term course of pulmonary arterial pressure in chronic obstructive pulmonary disease. Am Rev Respir Dis 1984; 130:993-998

(24) Weitzenblum E, Schrijen F, Mohan-Kumar T, et al. Variability of the pulmonary vascular response to acute hypoxia in chronic bronchitis. Chest 1988; 94:772-778

(25) Seals DR, Johnson DG, Fregosi RF. Hyperoxia lowers sympathetic activity at rest but not during exercise in humans. Am J Physiol 1991; 260:R873-R878

(26) Lund VE, Kentala E, Scheinin H, et al. Heart rate variability in healthy volunteers during normobaric and hyperbaric hyperoxia. Acta Physiol Scand 1999; 167:29-35

* From the Departments of Pulmonary Medicine (Drs. Smit, Vonk-Noordegraaf, Postmus, de Vries, and Boonstra, and Ms. van der Weijden) and Clinical Physics and Informatics (Dr. Marcus), Vrije Universiteit Medical Center, Amsterdam, the Netherlands. Manuscript received January 2, 2002; revision accepted November 27, 2002.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: permissions@chestnet.org).

Correspondence to: Henk J. Smit, MD, Vrije Universiteit Medical Center Department of Pulmonary Medicine, PO Box 7057 1007 MB Amsterdam, the Netherlands; e-mail: HJ. Smit@VUmc.nl

COPYRIGHT 2003 American College of Chest Physicians
COPYRIGHT 2003 Gale Group