Introduction

Appropriate gas exchange at the lowest risk of ventilator-induced lung injury is the general aim of ventilator therapy. The choice of the optimum ventilation strategy is controversial, especially in patients with acute lung injury (ALI) or with adult or infant respiratory distress syndrome [1, 2, 3]. Unfortunately, at present it is not possible to monitor the instantaneous state of regional lung function at the bedside nor to evaluate the local effects of changes in ventilator strategy or other therapeutic interventions.

Electrical impedance tomography (EIT) is a new method for monitoring regional lung function that has been proposed as a tool for optimising ventilatory management [4, 5]. An approach to determine the optimum ventilatory pressures by EIT measurement of regional lung volume changes during a static stepwise pressure-volume manoeuvre was recently reported [6]. This is an important step towards the future clinical use of EIT, although the low EIT acquisition rate in that study did not permit dynamic assessment of distribution of ventilation. Modern EIT technology is expected to assess not only the static but also the dynamic lung volume changes.

The aim of this brief report is to demonstrate the monitoring capacity of up-to-date EIT technology regarding the temporal and spatial distribution of lung aeration and ventilation in an experimental model of ALI.

Materials and methods

The experimental protocol adhered to the guidelines on animal experimentation (“Principles of laboratory animal care”, NIH publication no. 86-23, revised 1985) and was approved by the state animal care committee. A newborn piglet (body weight: 2 kg) was sedated with azaperon (2 mg/kg). Anaesthesia was induced with ketamine (35 mg/kg) and maintained with fentanyl (20 µg/kg per hour) and midazolam (0.3 mg/kg per hour). The animal was tracheotomised and mechanically ventilated, and intravenous and intra-arterial catheters were inserted. Muscle paralysis was induced and maintained with pancuronium (0.3 mg/kg per hour).

ALI was induced by repeated bronchoalveolar lavage with warm saline (50 ml/kg). A total of ten lavages were performed, with the desired end-point being arterial PO2 lower than 100 mmHg in FIO2 1.0. The animal was turned prone midway through the lavage sequence and returned to the supine position at its end.

Sixteen ECG electrodes were placed around the circumference of the chest at the level of the 6th parasternal intercostal space and connected with the EIT device (GoeMF II system, Department of Anaesthesiological Research, University of Göttingen) [7]. Raw EIT data generated by repeated rotating pulses of electrical current (5 mA rms, 50 kHz) and measurements of surface voltages between adjacent pairs of electrodes, were acquired and transformed into EIT scans at a rate of 13 scans/s using a back-projection reconstruction procedure [8].

The first EIT measurement was performed during the following ventilatory manoeuvre: Following a 30-s disconnection from the ventilator the piglet was re-ventilated in volume-controlled mode with a tidal volume (VT) of 10 ml/kg at 40 breaths/min. Over a period of 120 s PEEP was increased from 0 to 30 cmH2O in 5 cmH2O increments (stepwise inflation) and thereafter returned in 5 cmH2O steps back to baseline (stepwise deflation). Thirty minutes after intratracheal administration of surfactant (120 mg/kg; Curosurf, Nycomed, Munich, Germany), a second measurement was performed during a PEEP manoeuvre identical with the first one. A total of 1500 EIT scans were collected during each measurement.

In subsequent off-line analysis, functional EIT scans showing the instantaneous distribution of regional lung aeration and ventilation during each incremental and decremental PEEP step were generated. The influence of PEEP on regional lung aeration was evaluated with scans showing the shift in local end-expiratory lung volume (EELV) at each PEEP step in comparison with the lung volume during the apnoeic phase preceding each manoeuvre. The calculation was performed by subtracting the individual pixel values of relative impedance change during the apnoeic phase from the corresponding end-expiratory values of the last breath during each PEEP step. Regional lung ventilation at each PEEP level was visualised by scans showing the instantaneous distribution of regional VT. To determine regional VT local differences between end-inspiratory and end-expiratory values of relative impedance change were calculated from the last breaths of all PEEP steps. Both evaluation procedures are described in detail in [9]. Finally, tracings of local impedance change, revealing the dynamics of regional aeration and VT changes, were plotted in selected regions of interest (ROI).

Results

The changes in regional lung aeration occurring during the manouevres are shown in Fig. 1. In the surfactant-depleted lung only the non-dependent lung regions showed an increase in aeration during the initial PEEP increments, whereas after surfactant treatment a more uniform increase in aeration was noted throughout the lung fields. At the highest lung volume (30 cmH2O of PEEP) the pattern of distribution was almost identical in the surfactant-depleted and treated lung. Greater and more homogeneous aeration was found at identical PEEP during stepwise deflation, both before and after surfactant treatment, than during inflation. Regional collapse was noted earlier in ALI (see the disappearance of light areas at 5 cmH2O of PEEP).

Fig. 1
figure 1

Changes in regional lung aeration at end-expiration in the chest cross-section of a piglet during volume-controlled ventilation at different positive end-expiratory pressure levels in comparison with the gas content during an apnoeic phase preceding the trial. The electrical impedance tomography (EIT) scans were obtained after induction of acute lung injury (top) and subsequent surfactant treatment (bottom). The higher the regional increase in aeration the lighter the area in the EIT scan. The scans were generated with the following orientation: top anterior, left right part of the chest

Figure 2 shows the distribution of VT at all PEEP levels before and after surfactant administration. In the surfactant-depleted animal only the ventral lung regions were initially ventilated. The ventilation of dependent lung regions improved with each PEEP increment, such that the whole lung was ventilated relatively uniformly at 20 cmH2O of PEEP. Homogeneous ventilation distribution was initially preserved during stepwise deflation but became increasingly uneven below 10 cmH2O of PEEP. Return to the baseline was associated with complete loss of ventilation to the dependent region of the right lung. Ventilation in the corresponding region on the left was maintained.

Fig. 2
figure 2

Distribution of tidal volume in the chest cross-section of a piglet with induced acute lavage lung injury (top) and after surfactant treatment (bottom) during volume-controlled ventilation at different PEEP levels. The higher the regional VT the lighter the area in the EIT scan

After surfactant treatment preferential ventilation of the right ventral lung region indicative of inhomogeneous effect and/or administration of surfactant was found at a PEEP of 0 cmH2O, both before and after the manoeuvre. In contrast with the surfactant-depleted lung, an earlier distribution of ventilation to the whole lung was observed during stepwise inflation. The ventilation distribution remained homogeneous with PEEP decrements, although slightly more pronounced ventilation of the left lower lung region was identified at 10 and 5 cmH2O of PEEP.

Figure 3 shows the tracings of instantaneous impedance change in four ROIs in the right lung during the PEEP manoeuvres. The tracings allow an assessment of the immediate effect of PEEP adjustment on local EELV and VT, a comparison of regions lying at different vertical heights, and an estimation of the effect of surfactant treatment. The following observations can be made: (a) In the uppermost region the pre- and postsurfactant courses of EELV and VT changes were similar. Local VT in this region was higher at low PEEP than at high, reflecting preferential ventilation of the ventral lung units at low lung volume, both before and after surfactant therapy. During stepwise deflation a transient rise in EELV was noted after the first PEEP decreases, suggesting that there had been regional compression related to overdistension of adjacent lung units at the highest PEEP settings. A falling EELV, indicative of lung collapse, was present during the late steps in ALI. (b) In the lowermost region the courses of EELV and VT changes were dissimilar during stepwise inflation and deflation. No gas entered this region during the first increments in PEEP, with ventilation occurring for the first time at a PEEP of 20 cmH2O and 10 cmH2O before and after surfactant treatment respectively. (c) In all lung regions, during stepwise deflation volume was considerably higher at each PEEP decrement in the surfactant-treated lung, indicating higher mechanical stability of the lung tissue. A volume gain existed in each ROI at the end of the trial after surfactant treatment.

Fig. 3
figure 3

Tracings of local relative impedance change (right top, dark thick lines) during the PEEP manoeuvre in four regions of interest in the right lung (left top) before and after surfactant treatment. An increase in local aeration is accompanied by an increase in electrical impedance; the small fluctuations in the impedance signal represent the individual breaths. For better comparison and identification of instantaneous changes of end-expiratory lung volume and tidal volume with PEEP the individual tracings were normalised and plotted together with the tracing of average relative impedance change in the whole thoracic cross-section (light thin lines). Timing of the PEEP manoeuvre (left bottom) and the tracings of airway pressure (right bottom) are shown. Z impedance; Pao pressure at the airway opening

Discussion

EIT is an emerging monitoring technique capable of detecting local changes in pulmonary air content [10]. To establish EIT in the clinical environment adequate data acquisition and evaluation procedures suitable for routine use must be developed. Recently an approach to define appropriate ventilatory pressures by EIT measurements of lung volume changes during a pressure-volume manoeuvre under static conditions has been reported [6]. In that study the acquisition time for each EIT scan was 1.13 s, and the data had to be averaged to achieve acceptable signal-to-noise ratio. Given that lung collapse and recruitment often occur with time constants of less than 1 s in ALI [11, 12], the monitoring capacity of EIT may significantly be enhanced by higher scanning rates and improved data quality.

The latest technological advances [7] enable scanning rates up to 44 Hz. This means that information on the temporal and spatial distribution of lung aeration and ventilation can be obtained in a dynamic sense. This brief report demonstrates how EIT is able to track the instantaneous response of lung tissue to a simple recruitment-derecruitment manoeuvre and to document the effect of surfactant administration. The EIT scans showing shifts in EELV and VT distribution, as well as the tracings of EIT data in different ROIs, make it clear that in spite of its limited spatial resolution EIT can provide valuable information regarding the state of aeration and ventilation of regions within the lung. These issues are crucial for the definition of adequate ventilator strategies to allow lung-protective ventilation in injured lungs [13, 14, 15].

We hope that our report will stimulate further research aimed at applying EIT as a tool for optimising ventilatory management.