Mechanical ventilation is a vital supportive measure that allows the maintenance of alveolar ventilation and blood oxygenation during both routine anaesthesia and the care of the critically ill. However, mechanical stresses during positive pressure ventilation may also have deleterious pulmonary consequences, leading to inflammation and adverse structural and functional changes,1 as well as haemodynamic instability from impeded venous return.2 These factors can contribute significantly to respiratory morbidity and mortality in mechanically ventilated patients. Finding the best tidal volume and pressure to maintain adequate gas exchange during mechanical ventilation, while minimising mechanical stress to the lung, remains a major challenge. Traditional ventilator modes allow the setting of a desired tidal volume or peak inspiratory pressure (PIP). Modern mechanical ventilators provide more sophisticated ventilation modes such as pressure-regulated volume control (PRVC), an assist/control mode of ventilation. In this mode, the delivery of tidal volume is guaranteed without exceeding a preset inspiratory pressure limit by decelerating inspiratory flow and modifying inspiratory time on a cycle to cycle basis.
Previous studies have assessed the differences between PRVC and conventional volume control (VC) ventilation in both infants and adults.3–5 These studies suggest that PRVC ventilation allows a reduction in PIPs without disturbing cardiac output, airway pressures and gas exchange.5 It is thought that the rapid early gas delivery in this ventilation mode allows raised and sustained respiratory pressure for a longer time during the respiratory cycle, thereby enhancing lung recruitment and ventilation distribution.6 This assumption, however, is not based on direct evidence, and currently, it is not understood how varying flow patterns might alter the regional distribution of ventilation within the lung. To our knowledge, the few laboratory and clinical studies comparing lung function between different ventilation modes have not addressed the question of regional functional and structural differences within the lung.4,7
We have developed a xenon-enhanced computed tomography (CT) technique that uses synchrotron-generated X-rays to simultaneously image lung morphology and tissue density, in addition to the regional distribution of ventilation.8,9 We have previously shown that combining data on regional lung aeration and regional ventilation allows the definition of a spectrum of regional mechanical behaviour ranging from tidal recruitment to complete air trapping.8 Application of this functional imaging technique in combination with the assessment of airway and respiratory tissue mechanics allows a detailed assessment of the conducting airways and alveolar compartments under different ventilation modes.
The goal of the present study was, therefore, to compare the effects of PRVC and VC modes on the distribution of regional lung aeration and ventilation using synchrotron radiation imaging and to assess the forced oscillatory mechanics of the respiratory system with both ventilation modes in normal lung and after inducing lung injury by lavage in anesthetised rabbits. Our main hypothesis was that a decelerating flow pattern would allow a reduction in PIPs and also a more even distribution of regional lung ventilation.
Materials and methods
Animal care and the performance of the experiments were in accordance with the Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes. The Internal Evaluation Committee for Animal Welfare in Research and the Safety Group of the European Synchrotron Radiation Facility approved this experiment MD−679 on 12 March 2012.
The experiments were performed on eight male New Zealand white rabbits (2.9 ± 0.1 kg). Anaesthesia was induced by intravenous injection of thiopental sodium (25 mg kg−1) via a 22-G catheter introduced into the marginal ear vein under local anaesthesia (5% topical lidocaine). Tracheotomy was performed with a no. 3 Portex tube (Smiths Medical, Kent, UK), and the lungs were mechanically ventilated by a commercial neonatal ventilator (Servo-I, Maquet Critical Care, Solna, Sweden) with an electronic modification that allowed the synchronisation of mechanical ventilation with the image acquisition.
The left carotid artery and jugular vein were catheterised for blood gas measurements and drug delivery. Anaesthesia was then maintained with 0.1 mg kg−1 h−1 of intravenous midazolam, and analgesia by intravenous administration of fentanyl (50 μg kg−1 h−1). When vital signs indicated adequate anaesthesia, continuous intravenous atracurium (1.0 mg kg−1 h−1) was infused. The animal was immobilised in a custom-made plastic holder in the vertical position for imaging.
Synchrotron radiation computed tomography imaging
The experiments were performed at the Biomedical Beamline of the European Synchrotron Radiation Facility (Grenoble, France). The K-edge subtraction (KES) imaging technique was used as described previously.8,10–12 This technique allows quantitative measurements of regional specific ventilation (
) as well as lung tissue density within the same images simultaneously. The technique uses two X-ray beams tuned at slightly different energies above and below the Xe K-edge (34.56 keV), the binding energy of the K shell electrons. X-rays from a synchrotron radiation source are required since, as opposed to standard X-ray sources, they allow the selection of monochromatic beams from the full X-ray spectrum while conserving enough intensity for imaging with sufficient temporal resolution. Two CT images are thus simultaneously acquired during the inhalation of stable Xe (20%) in air. The densities of tissue and Xe can be separately calculated in each image voxel using a specifically developed computer algorithm11 explained in detail elsewhere.12 The ‘Xe-density’ image allows the direct quantitative measurement of this gas within the airspaces, and that of the regional gas volume. Dynamic KES imaging during Xe wash-in or wash-out allows the measurement of regional specific ventilation (
).9 This is performed by acquiring a series of 10 to 12 KES images with a set number of respiratory cycles in between. A ‘tissue-density’ image obtained from the same data allows the assessment of lung morphology and quantitative measurement of the regional lung aeration.8 The horizontal pixel size of the images was 0.35 mm, and the height of the beams was 0.7 mm at the animal position.
Images were processed by using the MatLab programming package (Mathworks Inc., Natick, Massachusetts, USA).8 Lung tissue was selected within the tissue-density CT images by region-growing segmentation, with manual editing in injured lung to avoid errors due to increased lung density. The local specific ventilation, defined as ventilation normalised to the gas volume within the voxel (
), was calculated from the time constant of the Xe wash-in using a single compartment model fit of Xe concentration vs. time. To minimise the effects of statistical truncations and image registration errors, Xe-density images were smoothed with a 5 × 5 pixel moving average window prior to the model fit. In each
image, the histogram of
was calculated, and a log-normal function applied. The median (μ) and SD (σ) of the distribution were extracted from the baseline fit.13
values were categorised as follows: ‘no ventilation’, defined as:
less than 0.5 min−1; Low
from 0.5 min−1 to μ − 2 σ; Normal
. Trapping was defined as aerated lung areas with no
. The area of lung within the images was computed and totalled over the four axial image slices to calculate the ‘total lung region of interest (ROI) area’ and the lung tissue density (D) in mg cm−3 was converted to Hounsfield units10 (HU). Hyperinflation was defined as a lung density below −900 HU.14 Lung regions with a density of −900 to 500 HU were qualified as normally aerated, regions with density of −500 to −100 HU as poorly aerated, and nonaeration was defined as lung regions with a density from −100 HU to 0 HU, based on previously published studies.14,15 To characterise the functional behaviour of normally aerated, poorly aerated and hyperinflated lung regions, the areas were divided into sub-categories based on corresponding ventilation categories as described earlier. Comparison of lung aeration and
was performed pixel by pixel, and each sub-category was expressed as percentage of the total lung ROI area within the image slice.
Regional dynamic strain maps were calculated based on specific ventilation. Dynamic strain1 (not including the static strain induced by PEEP) within a given image voxel was defined as:
Where Vgvox is the end-expiratory volume of gas contained in the voxel and dVvox, the tidal volume per voxel. Vgvox was computed based on the tissue-density images as described above. The dVvox is given by:
RR is the respiratory rate during the Xe wash-in manoeuvre. Combining the above equations gives:
Measurement of respiratory mechanics
The airway and respiratory tissue variables were assessed by using the forced oscillation technique at low frequencies. These measurements were achieved by introducing a loudspeaker-generated small-amplitude (1 cmH2O peak to peak) pressure forcing signal (0.5 to 21 Hz) into the trachea via a polyethylene tube (100 cm length, 0.375 cm internal diameter), whereas the mechanical ventilation was paused at end-expiration. The pressure inside the loudspeaker chamber was maintained at the level of PEEP to maintain pressure constant during the recordings. Lateral pressures were measured at the loudspeaker end (P1) and the tracheal end (P2) of the wave-tube with miniature pressure transducers (ICS 33NA00D, Malpitas, California, USA). These pressure signals were low-pass filtered (corner frequency of 25 Hz) and digitised at a sampling rate of 128 Hz. The pressure transfer function (P1/P2) was calculated by fast Fourier transformation from the 8-s recordings and the input impedance of the respiratory system (Zrs) was computed from this pressure transfer function as the load impedance of the wave tube.16 Three to five Zrs spectra were ensemble-averaged under each experimental condition. A model that includes airway resistance (Raw), inertance (Iaw) in series with constant-phase tissue compartments incorporating tissue damping (G) and elastance (H) was fitted to the averaged Zrs data.17
At the onset of mechanical ventilation, VC or PRVC ventilation mode was initiated in a randomised order. The following settings were used in both ventilation modes: tidal volume: 7 ml kg−1; respiratory rate: 40 min−1; PEEP: 3 cmH2O; inspired oxygen fraction (FiO2): 0.5. These ventilator settings resulted in an end-tidal CO2 of 5.5 to 6 kPa. After reaching steady-state conditions in systemic haemodynamic and ventilation variables, a recruitment manoeuvre was performed by inflating the lung to a peak pressure of 30 cmH2O, to standardise the volume history. The lungs were then ventilated for 10 min, and a set of Zrs recordings was collected during short end-expiratory pauses followed by acquisition of 12 subsequent KES images during Xe wash-in at four approximately equidistant axial positions from the apical (nondependent) to the caudal (dependent) lung. The axial positions were standardised based on the apex-diaphragm distance on a thoracic projection image. The ventilation mode was then switched and the same measurement sequence was repeated. After the baseline measurements were completed, whole lung lavages were performed by instilling 0.9% saline into the endotracheal cannula at 37°C. Gentle manual suctioning was employed to facilitate lavage fluid withdrawal. The ventilation was resumed for 2 min and this procedure was repeated. A total volume of 100 ml kg−1 was instilled over five sequential lavages. After whole lung lavage, a recruitment manoeuvre was performed as described above, and the respiratory mechanical measurements and the imaging acquisitions were repeated in the same manner as in the baseline condition.
The scatters of the variables were expressed by the SEM values, except for blood gas data where scatter was expressed as interquartile range. The sample size was estimated assuming a power of 0.8, α of 0.05, and an effect size of 20% for Raw. The Shapiro–Wilk test was used to test data for normality. Both the mechanical and imaging values were normally distributed. Two-way repeated measures analysis of variance was applied to evaluate the effects of the variables on the mechanical and imaging values with ventilation mode (VC vs. PRVC) and experimental condition (baseline vs. lavage) as within-subject variables. Pairwise comparisons were performed by using Holm–Sidak multiple comparison procedures. Pearson correlation tests were performed to test the statistical significance of the relationships between the imaging and respiratory mechanical variables. The statistical analyses were conducted by SigmaPlot (version 12.5, Systat Software, Inc., Chicago, Illinois, USA). Statistical tests were carried out with the significance level set at P < 0.05.
Effect of ventilation mode on airway pressures
The effect of ventilation mode on the ventilation pressures at baseline and following lung lavage are shown in Fig. 1. Delivering the same tidal volume with the PRVC mode resulted in significantly lower PIP in the healthy lungs (−11.1 ± 2.5%, P < 0.001). Whole-lung lavage significantly elevated PIP (P < 0.001) but it remained lower with the PRVC mode −12.7 ± 1.7%, P < 0.001. The airflow pattern during inspiration was different between the two ventilation modes. Airflow decelerated during inspiration on the PRVC mode as shown in a representative tracing in Fig. 2. The arterial blood gases are summarised in Table 1. There were no significant differences in the gas exchange indices between the two ventilation modes. Whole-lung lavage adversely affected gas exchange as reflected by a decrease in PaO2 and an increase in PaCO2 with no detectable alterations in the acid–basis status.
Effect of ventilation mode on respiratory mechanics
The airway and respiratory tissue mechanical variables obtained under VC and PRVC ventilation modes are shown in Fig. 3. Lung lavage led to a significant deterioration in tissue mechanics independent of the mode of ventilation, with significant increases in G for VC (P = 0.02) and PRVC (P = 0.03), and in H (P = 0.006 and P = 0.001, respectively). The Raw did not change significantly, either under VC (P = 0.089) or PRVC (P = 0.054). Overall, there was no evidence for a difference in respiratory mechanics between the two ventilator modes.
Effect of ventilation modes on regional lung function
Figure 4 demonstrates sample KES ventilation images, tissue-density images showing the distribution of
and maps depicting the regional mechanical behaviour of the lung periphery, based on combined aeration and regional ventilation data, in a representative rabbit. Whole-lung lavage led to the appearance of patchy lung regions with varying degrees of poor aeration, down to complete lack of aeration. This heterogeneous deterioration of the regional ventilation resulted in a redistribution of ventilation to the remaining normally aerated lung regions, which were then unevenly hyperventilated. In contrast with the clearly notable effects of lung lavage, no major difference in regional lung ventilation distribution was apparent between the ventilation modes either at baseline or following lavage.
The relative area of lung regions in each of the categories defined based on aeration and
are summarised in Fig. 5. Following lavage, the relative area of poorly aerated lung increased significantly. Part of the poorly aerated regions had faster specific ventilation. However, a significant portion of the poorly aerated regions had either normal or even reduced
. The relative amount of nonaerated lung regions significantly increased also. Neither lavage nor ventilation mode had a significant effect on the amount of air trapping. The area of lung regions that were hyperinflated at baseline was significantly reduced after lavage. Finally, in the remaining normally aerated lung regions,
was significantly increased in a subset of regions because of ventilation redistribution from nonventilated and poorly ventilated areas.
The computed maps of regional lung strain are shown in Fig. 6. The mean values of regional strain per axial image slice were significantly increased after whole-lung lavage (0.125 ± 0.03 vs. 0.109 ± 0.02 on VC, P = 0.025; 0.119 ± 0.03 vs. 0.100 ± 0.02 on PRVC; P = 0.007). However, mean regional lung strain was not significantly different between the two ventilation modes, either at baseline (P = 0.197), or following lung lavage (P = 0.417).
Correlation between regional lung function indices and global respiratory mechanics
The relationship between indices reflecting the area of lung regions in each functional category and the respiratory tissue mechanics are demonstrated on Fig. 7. There was a statistically significant correlation between the amount of nonaerated and poorly aerated lung regions and the magnitude of G and H (P < 0.001 for both). The correlations between the amount of hyperinflated areas and the tissue mechanical variables were somewhat weaker, but still statistically significant (P = 0.02 and 0.007 for G and H, respectively).
Detailed analyses of the lung functional differences between VC and PRVC ventilation modes in the present study have shown that both led to comparable regional lung ventilation distribution, and also airway and respiratory mechanics. We have previously shown that the regional mechanical behaviour of the peripheral lung units could be described by combining data on regional aeration and specific ventilation.8 In the present study, the description of regional lung function using this approach demonstrated no differences between the two ventilation modes in healthy lungs. The similarities in regional lung function between the two ventilation modes were manifested in identical gas exchange values with both ventilation strategies in both healthy and injured lungs, despite the lower peak pressures generated by the PRVC mode.
Optimisation of mechanical ventilation to provide adequate gas exchange while maintaining minimal lung inflation pressures is a major challenge in anaesthesia and intensive care, particularly in injured lungs. New ventilation modes have been developed to meet this demand using protective and adaptive strategies to manage continuous changes in respiratory resistance and compliance. In this regard, PRVC mode claims to guarantee the preset tidal volume while meeting the gas exchange requirements, by decelerating the inspiratory flow to minimise positive lung inflation pressure (Fig. 2). The results of the present study confirmed the delivery of the guaranteed tidal volume with lower peak airway pressures. The lower peak pressure can be attributed to a lower resistive pressure at end-inspiration due to the decelerating flow on PRVC mode. Because lung elastance and tidal volumes were not different, the elastic pressures were similar between the two ventilation modes.18,19 Previous clinical studies have suggested potential benefits from a reduction in PIP in neonates with respiratory distress syndrome, particularly in lowering the incidence of complications resulting from mechanical ventilation.20 However, clinical evidence of the positive effect of PRVC ventilation on outcomes such as respiratory mechanics, gas exchange or haemodynamic status is yet lacking.21
The model adopted in the present study features a highly heterogeneous collapsible lung with the development of poorly aerated and nonaerated regions,22–25 intended to mimic a key mechanical behaviour of acute respiratory distress syndrome.8,26 Under these conditions, the changes in the respiratory mechanical and regional lung ventilation variables following lung lavage agreed with those observed previously under similar experimental conditions.8 The results of the present study allowed us not only to evaluate the lavage-induced changes in the indices related to respiratory mechanics and regional ventilation, but also to assess the relationships between these outcomes under the two ventilation modes. As far as we are aware, this is the first study to compare airway and tissue mechanical indices between VC and PRVC ventilation modes.
The most remarkable finding of this study is that the regional ventilation distribution and the mechanical behaviour of the lung with PRVC ventilation were comparable with those obtained under constant inspiratory flow, in both healthy and injured lungs. The few available studies comparing the PRVC and VC modes failed to demonstrate benefit from the protective approach of the former and even pointed to potential deleterious effects of PRVC.27–29 However, all of the previous studies used relatively high tidal volumes, generating both high peak inspiratory flows and lung overdistension, which may limit the interpretation of their data in the context of a protective ventilation strategy. Therefore, in line with the recent guidelines on protective ventilation strategy in clinical practice,30 in the present study, we targeted a relatively low tidal volume (7 ml kg−1) with a moderate PEEP of 3 cmH2O. Under these conditions, there was no evidence of any major adverse effects of PRVC mode on global lung functional or regional lung ventilation, even in the presence of lung injury promoting alveolar derecruitment.
Specific ventilation describes the rate at which Xe washes into the acini in a given lung region. This variable is, therefore, determined by both the global minute ventilation and the local mechanical time-constant of a given lung region. It follows that
is increased by a reduced gas fraction within the acinar units in an imaged lung region. Conversely, a ‘normal’ or low
in a poorly aerated region suggests increased resistance of the subtending bronchi, either through narrowing or because of intermittent closure,8,31 whereas an increase in
in normally aerated lung regions implies that these zones receive a larger share of the tidal ventilation. In the presence of highly heterogeneous lung collapse induced by lavage, functional imaging revealed a significant redistribution of regional lung ventilation from the poorly aerated areas to zones with normal aeration, which had high specific ventilations (Fig. 5). Furthermore, a significant portion of the lung, mostly at the boundary of nonaerated regions (Fig. 4), was poorly aerated but showed high specific ventilation. This phenomenon can be interpreted as a reduction in the number of aerated alveoli within these lung units, which show shorter time-constants of xenon wash-in. Our imaging data showed the development of poorly and nonaerated lung regions under both ventilation modes. These phenomena are significant, as they promote mechanical stresses both within nonaerated lung units, and within normally aerated lung units, receiving a larger share of tidal ventilation.8,32,33 In this study, applying a decelerating flow did not result in a more homogeneous ventilation distribution. Nor did the regional ventilation images show evidence that higher inspiratory pressure and flow at the beginning of inspiration resulted in measurable lung recruitment. Similarly, regional lung strain (Fig. 6), although significantly higher after lavage-induced lung injury, was not significantly reduced on PRVC ventilation.
The close correlation between oscillatory tissue mechanical and functional imaging suggests that changes in tissue elastance and damping can be a good surrogate for the assessment of the development of altered regional ventilation. The stronger correlations with the amount of nonaerated and poorly aerated zones suggest that they are more sensitive for detecting lung volume losses. The existence of a negative correlation of G and H with the hyperinflated lung areas probably reflects the increased elastance within these regions because of an increased surface tension induced by lavage. However, the fact that hyperinflation correlates positively with elastance in normal lung as shown previously,8 indicates that the changes in the respiratory tissue variables can only be clearly evaluated by concomitant measurements of the effective lung volume.34
In summary, simultaneous measurements of regional lung aeration and ventilation in a lavage-induced model of lung injury using synchrotron imaging, showed that a decelerating flow pattern (PRVC) resulted in equivalent regional ventilation distribution, respiratory mechanics and gas exchange, in both normal and mechanically heterogeneous lungs with, however, a significantly lower peak pressure on PRVC ventilation. Our data suggest that the lower PIP on PRVC was due to the decelerating flow pattern rather than the ventilation distribution.
Acknowledgements relating to this article
Assistance with the study: this work was performed at the European Synchrotron Radiation Facility, Biomedical Beamline-ID17, Grenoble, France. The authors would like to thank Christian Nemoz, Thierry Brochard, Herwig Recquart and Charlene Caloud for their assistance with the study.
Financial support and sponsorship: this work was supported by the European Regional Development Fund #REG08009 and by the Picardie Regional Council (Amiens, France), the Tampere Tuberculosis Foundation (Helsinki, Finland), the European Synchrotron Radiation Facility (Grenoble, France), Swiss National Science Foundation grant 32003B-143331 (Bern, Switzerland), the Academy of Finland (Helsinki, Finland, grant 126747), Hungarian Basic Scientific Research Grant OTKA K81179 (Budapest, Hungary) and the Department of Anaesthesiology Pharmacology and Intensive Care, University Hospitals of Geneva (Geneva, Switzerland).
Conflicts of interest: WH has received a research grant from Maquet, Solna, Sweden for travel arrangements and equipment.
Presentation: preliminary data from this study were presented as a poster presentation at the American Society of Anesthesiologists (ASA) 2013 annual meeting, 11–15 October, San Francisco.
Comment from the editor: WH is a Deputy Editor-in-Chief of the European Journal of Anaesthesiology.
1. Chiumello D, Carlesso E, Cadringher P, et al. Lung stress and strain during mechanical ventilation for acute respiratory distress syndrome. Am J Respir Crit Care Med
2. Slutsky AS, Ranieri VM. Ventilator-induced lung injury. N Engl J Med
3. Guldager H, Nielsen SL, Carl P, Soerensen MB. A comparison of volume control and pressure-regulated volume control ventilation in acute respiratory failure. Crit Care
4. D’Angio CT, Chess PR, Kovacs SJ, et al. Pressure-regulated volume control ventilation vs synchronized intermittent mandatory ventilation for very low-birth-weight infants: a randomized controlled trial. Arch Pediatr Adolesc Med
5. Kocis KC, Dekeon MK, Rosen HK, et al. Pressure-regulated volume control vs volume control ventilation in infants after surgery for congenital heart disease. Pediatr Cardiol
6. Kesecioglu J, Gultuna I, Pompe JC, et al. Assessment of ventilation inhomogeneity and gas exchange with volume controlled ventilation and pressure regulated volume controlled ventilation on pigs with surfactant depleted lungs. Adv Exp Med Biol
7. Sjostrand UH, Lichtwarck-Aschoff M, Nielsen JB, et al. Different ventilatory approaches to keep the lung open. Intensive Care Med
8. Bayat S, Porra L, Albu G, et al. Effect of positive end-expiratory pressure on regional ventilation distribution during mechanical ventilation after surfactant depletion. Anesthesiology
9. Bayat S, Strengell S, Porra L, et al. Methacholine and ovalbumin challenges assessed by forced oscillations and synchrotron lung imaging. Am J Respir Crit Care Med
10. Hounsfield GN. Computed medical imaging. Science
11. Bayat S, Le Duc G, Porra L, et al. Quantitative functional lung imaging with synchrotron radiation using inhaled xenon as contrast agent. Phys Med Biol
12. Porra L, Suhonen H, Suortti P, et al. Effect of positive end-expiratory pressure on regional ventilation distribution during bronchoconstriction in rabbit studied by synchrotron radiation imaging. Crit Care Med
13. Porra L, Monfraix S, Berruyer G, et al. Effect of tidal volume on distribution of ventilation assessed by synchrotron radiation CT in rabbit. J Appl Physiol
14. Vieira SR, Puybasset L, Lu Q, et al. A scanographic assessment of pulmonary morphology in acute lung injury. Significance of the lower inflection point detected on the lung pressure-volume curve. Am J Respir Crit Care Med
15. Gattinoni L, Pesenti A, Avalli L, et al. Pressure-volume curve of total respiratory system in acute respiratory failure. Computed tomographic scan study. Am Rev Respir Dis
16. Petak F, Hantos Z, Adamicza A, et al. Methacholine-induced bronchoconstriction in rats: effects of intravenous vs. aerosol delivery. J Appl Physiol
17. Hantos Z, Daroczy B, Suki B, et al. Input impedance and peripheral inhomogeneity of dog lungs. J Appl Physiol
18. Alvarez A, Subirana M, Benito S. Decelerating flow ventilation effects in acute respiratory failure. J Crit Care
19. Ferrando C, Garcia M, Gutierrez A, et al. Effects of different flow patterns and end-inspiratory pause on oxygenation and ventilation in newborn piglets: an experimental study. BMC Anesthesiol
20. Piotrowski A, Sobala W, Kawczynski P. Patient-initiated, pressure-regulated, volume-controlled ventilation compared with intermittent mandatory ventilation in neonates: a prospective, randomised study. Intensive Care Med
21. Rittayamai N, Katsios CM, Beloncle F, et al. Pressure-controlled vs volume-controlled ventilation in acute respiratory failure: a physiology-based narrative and systematic review. Chest
22. Habre W, Scalfaro P, Schutz N, et al. Measuring end-expiratory lung volume and pulmonary mechanics to detect early lung function impairment in rabbits. Respir Physiol Neurobiol
23. Bellardine Black CL, Hoffman AM, Tsai LW, et al. Relationship between dynamic respiratory mechanics and disease heterogeneity in sheep lavage injury. Crit Care Med
24. David M, Karmrodt J, Bletz C, et al. Analysis of atelectasis, ventilated, and hyperinflated lung during mechanical ventilation by dynamic CT. Chest
25. Rudolph A, Markstaller K, Gast KK, et al. Visualization of alveolar recruitment in a porcine model of unilateral lung lavage using 3He-MRI. Acta Anaesthesiol Scand
26. Ware LB, Matthay MA. The acute respiratory distress syndrome. N Engl J Med
27. Roth H, Luecke T, Deventer B, et al. Pulmonary gas distribution during ventilation with different inspiratory flow patterns in experimental lung injury: a computed tomography study. Acta Anaesthesiol Scand
28. Dembinski R, Henzler D, Bensberg R, et al. Ventilation-perfusion distribution related to different inspiratory flow patterns in experimental lung injury. Anesth Analg
29. Maeda Y, Fujino Y, Uchiyama A, et al. Effects of peak inspiratory flow on development of ventilator-induced lung injury in rabbits. Anesthesiology
30. Dellinger RP, Levy MM, Carlet JM, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med
31. Wellman TJ, Winkler T, Costa EL, et al. Effect of regional lung inflation on ventilation heterogeneity at different length-scales during mechanical ventilation of normal sheep lungs. J Appl Physiol
32. Mead J, Takishima T, Leith D. Stress distribution in lungs: a model of pulmonary elasticity. J Appl Physiol
33. Mertens M, Tabuchi A, Meissner S, et al. Alveolar dynamics in acute lung injury: heterogeneous distension rather than cyclic opening and collapse. Crit Care Med
34. Albu G, Wallin M, Hallback M, et al. Comparison of static end-expiratory and effective lung volumes for gas exchange in healthy and surfactant-depleted lungs. Anesthesiology