Continuous non-invasive monitoring of tidal volumes by measurement of tidal impedance in neonatal piglets.

Abstract

Electrical Impedance measurements can be used to estimate the content of intra-thoracic air and thereby give information on pulmonary ventilation. Conventional Impedance measurements mainly indicate relative changes, but no information concerning air-volume is given. The study was performed to test whether a 3-point-calibration with known tidal volumes (VT) during conventional mechanical ventilation (CMV) allows subsequent calculation of VT from total Tidal-Impedance (tTI) measurements using Quadrant Impedance Measurement (QIM). In addition the distribution of TI in different regions of the thorax was examined.
QIM was performed in five neonatal piglets during volume-controlled CMV. tTI values at three different VT (4, 6, 8 ml/kg) were used to establish individual calibration curves. Subsequently, each animal was ventilated with different patterns of varying VT (2-10 ml/kg) at different PEEP levels (0, 3, 6, 9, 12 cmH(2)O). VT variation was repeated after surfactant depletion by bronchoalveolar lavage. VT was calculated from tTI values (VT(calc)) and compared to the VT delivered by the ventilator (VT(PNT)). Bland-Altman analysis revealed good agreement between VT(calc) and VT(PNT) before (bias -0.08 ml; limits of agreement -1.18 to 1.02 ml at PEEP = 3 cmH(2)O) and after surfactant depletion (bias -0.17 ml; limits of agreement -1.57 to 1.22 ml at PEEP = 3 cmH(2)O). At higher PEEP levels VT(calc) was lower than VT(PNT), when only one fixed calibration curve (at PEEP 3 cmH(2)O) was used. With a new calibration curve at each PEEP level the method showed similar accuracy at each PEEP level. TI showed a homogeneous distribution over the four assessed quadrants with a shift toward caudal regions of the thorax with increasing VT.
Tidal Impedance values could be used for precise and accurate calculation of VT during CMV in this animal study, when calibrated at each PEEP level.

Full text

Page 1

Continuous Non-Invasive Monitoring of Tidal Volumes
by Measurement of Tidal Impedance in Neonatal Piglets
Florian Kurth, Fabienne Zinnow, Alexandra Prakapenia, Sabrina Dietl, Stefan Winkler, Sascha Ifflaender,
Mario Ru ¨diger*, Wolfram Burkhardt
Department of Neonatology and Paediatric Intensive Care, University Hospital Carl Gustav Carus, Dresden, Germany
Abstract
Background: Electrical Impedance measurements can be used to estimate the content of intra-thoracic air and thereby give
information on pulmonary ventilation. Conventional Impedance measurements mainly indicate relative changes, but no
information concerning air-volume is given. The study was performed to test whether a 3-point-calibration with known tidal
volumes (VT) during conventional mechanical ventilation (CMV) allows subsequent calculation of VT from total Tidal-
Impedance (tTI) measurements using Quadrant Impedance Measurement (QIM). In addition the distribution of TI in different
regions of the thorax was examined.
Methodology and Principal Findings: QIM was performed in five neonatal piglets during volume-controlled CMV. tTI values
at three different VT (4, 6, 8 ml/kg) were used to establish individual calibration curves. Subsequently, each animal was
ventilated with different patterns of varying VT (2–10 ml/kg) at different PEEP levels (0, 3, 6, 9, 12 cmH2O). VT variation was
repeated after surfactant depletion by bronchoalveolar lavage. VT was calculated from tTI values (VTcalc) and compared to
the VT delivered by the ventilator (VTPNT). Bland-Altman analysis revealed good agreement between VTcalcand VTPNTbefore
(bias 20.08 ml; limits of agreement 21.18 to 1.02 ml at PEEP=3 cmH2O) and after surfactant depletion (bias 20.17 ml;
limits of agreement 21.57 to 1.22 ml at PEEP=3 cmH2O). At higher PEEP levels VTcalcwas lower than VTPNT, when only one
fixed calibration curve (at PEEP 3 cmH2O) was used. With a new calibration curve at each PEEP level the method showed
similar accuracy at each PEEP level. TI showed a homogeneous distribution over the four assessed quadrants with a shift
toward caudal regions of the thorax with increasing VT.
Conclusion: Tidal Impedance values could be used for precise and accurate calculation of VT during CMV in this animal
study, when calibrated at each PEEP level.
Citation: Kurth F, Zinnow F, Prakapenia A, Dietl S, Winkler S, et al. (2011) Continuous Non-Invasive Monitoring of Tidal Volumes by Measurement of Tidal
Impedance in Neonatal Piglets. PLoS ONE 6(6): e21003. doi:10.1371/journal.pone.0021003
Editor: Irwin Reiss, Erasmus University Rotterdam, Netherlands
Received December 31, 2010; Accepted May 17, 2011; Published June 7, 2011
Copyright: ? 2011 Kurth et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was supported by a grant of the Deutsche Forschungsgemeinschaft DFG (Ru1233/1-2). Florian Kurth holds a research grant of the Medical
Faculty of the Technical University Dresden (60.264). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of
the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: mario.ruediger@uniklinikum–dresden.de
Introduction
Measurements of lung volume changes are a prerequisite to
optimize mechanical respiratory support and study respiratory
mechanics. In order to avoid interference with the patient and the
respiratory support, these measurements should ideally be non
invasive and should not increase breathing load or dead space. In
addition to global data on tidal volume (VT) and functional
residual capacity (FRC), information on regional air distribution
would be of interest. In neonates the restrictions of the currently
available methods, such as interference with spontaneous breath-
ing, limit the possibility of reliable and practical bedside
monitoring of lung volumes in clinical routine [1,2].
Electrical Impedance measurement is a non-invasive, radiation-
free bedside technique which determines changes in electrical
conductivity of organs based on the variability in electrical
impedance between tissue, air and fluid. The technology has been
proposed for the estimation of lung volume changes more than 20
years ago [3] and has been widely employed for surveillance of the
respiratory frequency in the form of impedance pneumography in
patient monitors for a long time [4]. Moreover, impedance
measurements can be used to generate cross-sectional images of
impedance distribution within the thorax, a technique which is
known as electrical impedance tomography (EIT) and which has
gained heightened interest in respiratory and intensive care
recently [5].
Whereas EIT produces colour coded information on relative
impedance changes, it does not allow quantification of impedance
changes of the lung and expression as absolute volumes. Quadrant
Impedance Measurement (QIM) uses a different approach of
impedance measurement which virtually divides the thorax in 4
quadrants and determines impedance changes in these 4 regions.
The amplitudes of impedance changes during respiratory cycles
are thereby expressed as Tidal Impedance (TI).
The present animal study was performed to answer the
following questions: 1. How precisely and how accurately can
VT in millilitres be estimated by total Tidal Impedance (tTI i.e.
the sum of TI-values from all four quadrants) after three-point
PLoS ONE | www.plosone.org1 June 2011 | Volume 6 | Issue 6 | e21003

Page 2

calibration with known tidal volumes during conventional
mechanical ventilation? 2. Can one single calibration procedure
(at PEEP 3 cmH2O) be used for different PEEP levels or should
individual calibrations be performed? 3. How is the precision and
accuracy of the method influenced by lung injury as implemented
by saline lavage?
Materials and Methods
Animals
The study was performed in five anaesthetized newborn piglets
and all experiments were approved by the university and the state
committee for animal care and adhered to the national law on the
care and use of laboratory animals. Animals were treated as
previously described [6]. In short, Ketamin (20 mg/kg) and
Azaperon (5 mg/kg) were injected intramuscularly for pre-medica-
tion. Anaesthesia was induced by intravenous administration of
Morphine (2 mg/kg) and Midazolam (2 mg/kg) and subsequently
maintained with Morphine (20 mg/kg/hr), Midazolam (2 mg/kg/
hr) and Rocuronium (2 mg/kg/hr). An arterial line was placed in
the femoral artery, connected with a pressure transducer and kept
open with a continuous infusion of heparin-saline mixture (1 ml/
hr). Animals were intubated via tracheotomy (tubes of 3.5–4.5 mm
inner diameter with a side port, Vygon, Ecouen, France) and
ventilated in supine body position in a volume-controlled mode of
ventilation (Stephanie, Fa. Stephan, Gackenbach, Germany). Initial
ventilator settings were as follows: respiratory rate 60 breaths/min,
inspiratory time 0.4 sec, fractional concentration of oxygen in
inspired gas 1.0, tidal volume 6 ml/kg, positive end-expiratory
pressure (PEEP) 4 cmH2O. Monitoring of vital parameters was
performed using a neonatal patient monitor (HP, Palo Alto, CA,
USA). Blood gas analysis was performed once at every PEEP level
(ABL 500, Radiometer, Copenhagen, Denmark). Surfactant
depletion was implemented by repetitive bronchoalveolar with
warmed saline (30 ml/kg) as originally described by Lachmann et al.
[7]. The lavage procedure was repeated until PaO2remained below
13 kPa for at least 30 minutes.
Experimental Protocol
Stepwise variation of five different VT (2, 4, 6, 8, 10 ml/kg) was
performedin3 differentpatterns:1.Increasing/decreasing(2-4-6-8-
10-8-6-4-2 ml/kg), 2. Antipodal around VT 6 ml/kg (6-2-6-10-6-4-
6-8-6), 3. Random (using every one of the above indicated VT
values twice in random order). VT was kept constant for 30 seconds
at each step of variation. The identical variation was performed at 5
different PEEP levels (0, 3, 6, 9, 12 cmH2O), except for the random
part which was different at every PEEP level.
VT-variation was performed before and after surfactant
depletion following the identical protocol, resulting in 280 per
animal in total.
Measurements of VT were performed at the Y piece of the
ventilator using the fixed orifice pneumotachograph (PNT) of the
Stephanie ventilator (resistance 1.1 kPa/L/s at 5 L/min flow, dead
space 0.9 mL). The individual differential-pressure/flow character-
istic curve of the pneumotachograph was determined prior to the
experiments (Schaller Medizintechnik, Dresden, Germany) in order
to obtain the coefficients for a fourth grade polynomial model for
calculationof VT by the ventilator. Calibration of the PNT with the
built-in test-lung was carried out according to the manufacturer’s
instruction prior to each experiment [8–10].
Impedance Measurements
Impedance measurements were performed using the EIS/
Quadrant Impedance Monitoring (QIM device of EMS Biomed-
ical (EMS, Korneuburg, Austria). The device measures Tidal
Impedance(TI) from the amplitudes of thoracic impedance changes
during a ventilatory cycle. A constant current source which
generates a sinus-wave output current of 400 mA at a frequency of
5 KHz was used and current was applied via one ventral and one
dorsal current electrode. Voltage measurements for subsequent
determination of impedance were performed in four regions of the
thorax, cephalic left, cephalic right and caudal left and right at a
frequency of 500 Hz. For each region one ventral and one dorsal
electrode was placed according to a standardized placement
procedure (Figure 1). Disposable adhesive EEG electrodes were
used for both current injection and voltage measurements (Neuro-
line 720, Ambu, Bad Nauheim, Germany). Tidal Impedance in the
four quadrants was recorded separately via four channels and
expressed as quadrant Tidal Impedance (qTI) in Arbitrary Units
(AU). Total Tidal Impedance (tTI) was subsequently calculated as
sum of qTI from all four channels. A Bandpass filter with a lower
frequency of 0.2 Hz and an upper frequency of 1.5 Hz was used to
process incoming biosignals of the four channels. Data were
transferred to a PC via USB interface and analyzed with EIM client
4.0 software (EMS, Korneuburg, Austria).
Analysis of variability of TI
To investigate the within-subject variability of impedance
measurements during subsequent breathing cycles tTI values of
one random sample of 10 breathing cycles at a constant
VT=6 ml/kg at PEEP=3 cmH2O were assessed for each animal
and compared to the variability of volumes as measured by the
PNT at a constant VT=6 ml/kg. For subsequent calculation of
tTI the mean amplitude of 5 breathing cycles was used once VT
had stabilized at the respective step of variation.
In a second step, variability of impedance measurements during
repeated measurements was assessed using all tTI values of
corresponding VT levels during VT variation at one PEEP level.
Calibration procedure
Three point calibration of tTI was performed prior to VT
variation at each PEEP level in every animal, using three pre-
specified VT levels (4, 6, 8 ml/kg). Linear regression analysis was
performed in order to compute an equation which allowed
calculation of VT (VTcalc) from tTI based on the three tTI
measurements from known VT during calibration (VTcalc=inter-
cept+slope6tTI).
VT was calculated in two different ways: 1.using one calibration
at PEEP 3 cmH2O for all VT at all 5 PEEP levels. 2. Using the
individual calibration from each PEEP level for calculation of VT
at the respective PEEP level. Bland-Altman analysis was
performed in order to compare VTcalcwith VT as indicated from
the PNT (VTPNT) [11].
Analysis of distribution of TI in different quadrants
Analysis of distribution of TI was performed based on the
separate data from the 4 different quadrants (qTI). Therefore the
fraction of qTI divided by tTI was computed for all TI
measurements and expressed as fractional TI (fTI) for every
quadrant. Changes of fTI were assessed for different VT at
different PEEP levels.
Statistical analysis
Data of Impedance Measurements was exported to MS excel
and JMP (JMP 7.0, SAS Institute Inc., NC, USA) via EIM client
4.0 software data export. Bland Altman analysis was performed
using a plug-in for MS excel (analyze-it.com). Bias and Limits of
Tidal Volume Monitoring by Impedance Measurement
PLoS ONE | www.plosone.org2 June 2011 | Volume 6 | Issue 6 | e21003

Page 3

agreement were calculated from the means of corresponding VT
from individual piglets. Variability of tTI was expressed as
coefficient of variation (CV). Data are expressed as mean (6SD)
if normal distribution was assumed or median (range) in case no
normal distribution was assumed.
Results
Animals had a median (range) age of 8 (1–14) days and a
bodyweight of 1.9 (1.6–3.0) kg. All five animals remained stable
during the experimental period and no additional interventions
were necessary. A median (range) of 4 (3–7) bronchoalveolar
lavages were required to achieve stable post-lavage PaO2levels
below 13 kPa. Figure 2 depicts the courses of blood gases and
heart rate during the experiments. Continuous recording of
impedance measurements was possible in all animals and TI
measurements were available for all 280 VT variations in all
animals resulting in 1400 measurements in total. Figure 3 shows a
representative tracing signal from impedance measurements
during 5 breathing cycles.
Figure 1. Placement of electrodes. a) Schematic map of the piglet’s body and position of electrodes; b) detailed illustration of the sternum and
distances of the electrodes; the length of the sternum (s) is quartered. Cranial Right (CrR) and Cranial Left (CrL) ventral are set at the end of the first
quarter from the cranial end of the sternum right and left with the distance J s from the axis. The central electrode for current injection (CE) is set in
the middle (m) of the remaining L of the sternum. Caudal Right (CaR) and Caudal Left (CaL) are placed right and left from the caudal end of the
sternum with the distance between m and the caudal end of the sternum. Dorsal electrodes are set mirror-inverted to the ventral electrodes.
doi:10.1371/journal.pone.0021003.g001
Figure 2. Blood Gases and Heart Rate. Shown are mean (SD) values of blood gases and heart rate for n=5 animals. Blood gas analyses were
performed after measurements at one PEEP level before changing to the next PEEP level.
doi:10.1371/journal.pone.0021003.g002
Tidal Volume Monitoring by Impedance Measurement
PLoS ONE | www.plosone.org3 June 2011 | Volume 6 | Issue 6 | e21003

Page 4

Within-subject variability of impedance measurements
Assessment of tTI values from 10 consecutive breathing cycles
at VT=6 ml/kg in each animal showed median (range) CV of
0.08 (0.04–0.10) and 0.07 (0.03–0.07) pre- and post-lavage,
respectively. This variability was comparable to the variability of
VT between single breathing cycles as delivered by the ventilator
and measured by the pneumotachograph (0.06, range 0.05–0.09
and 0.06, range 0.04–0.09, pre- and post-lavage, respectively).
Median within-subject variability of repeated tTI measurements
of corresponding VT levels at different time points during
Figure 3. Tracing signal of impedance measurement during 5 breathing cycles. Shown are tracing signals of QIM measurements from 4
quadrants during 5 respiratory cycles. In addition flow and pressure curves from the ventilator are shown below in order to illustrate the time course
of impedance changes with respect to ventilatory cycles from the ventilator.
doi:10.1371/journal.pone.0021003.g003
Table 1. Within subject variability of repeated tTI measurements.
Prelavage Postlavage
Vt (ml/kg)2468102468 10
PEEP 00,10
(0,04–0,19)
0,08
(0,05–0,10)
0,06
(0,05–0,06)
0,04
(0,02–0,08)
0,03
(0,01–0,04)
0,08
(0,02–0,14)
0,04
(0,02–0,07)
0,05
(0,03–0,08)
0,03
(0,02–0,04)
0,02
(0,02–0,03)
PEEP 3 0,08
(0,05–0,15)
0,07
(0,01–0,10)
0,05
(0,03–0,06)
0,03
(0,02–0,04)
0,01
(0,01–0,06)
0,08
(0,07–0,22)
0,05
(0,04–0,08)
0,04
(0,03–0,10)
0,03
(0,02–0,09)
0,03
(0,02–0,06)
PEEP 6 0,09
(0,03–0,22)
0,04
(0,03–0,09)
0,04
(0,03–0,09)
0,03
(0,03–0,05)
0,02
(0,01–0,03)
0,06
(0,03–0,11)
0,05
(0,03–0,15)
0,04
(0,03–0,07)
0,02
(0,02–0,06)
0,02
(0,02–0,03)
PEEP 90,07
(0,04–0,37)
0,07
(0,03–0,09)
0,05
(0,02–0,08)
0,04
(0,02–0,04)
0,03
(0,01–0,04)
0,11
(0,07–0,18)
0,05
0,03–0,13)
0,05
(0,03–0,10)
0,03
(0,02–0,07)
0,03
(0,01–0,08)
PEEP 120,10
(0,04–0,13)
0,06
(0,02–0,10)
0,04
(0,03–0,06)
0,03
(0,02–0,06)
0,02
(0,01–0,04)
0,08
(0,03–0,20)
0,04
(0,02–0,07)
0,05
(0,04–0,06)
0,02
(0,01–0,08)
0,02
(0,02–0,03)
Shown are median (range) coefficients of variation (CV) of repeated tTI measurements for corresponding VT levels at different PEEP levels Pre- and Post-lavage for n=5
animals. Repeated measurements are taken at different time points during variation at one PEEP level.
doi:10.1371/journal.pone.0021003.t001
Tidal Volume Monitoring by Impedance Measurement
PLoS ONE | www.plosone.org4 June 2011 | Volume 6 | Issue 6 | e21003

Page 5

Table 2. Agreement between VTcalcand VTPNTin Bland-Altman analysis.
Pre-lavage Post-lavage
Calibration at
PEEP=3 cmH2O
Calibration at same
PEEP level
Calibration at
PEEP=3 cmH2O
Calibration at same
PEEP level
PEEP 0 cmH20 Bias (ml)
20.250.420.110.37
Limits of agreement (ml)
22.05–1.40 0.71–1.56
22.04–2.26
21.24–1.98
PEEP 3 cmH20 Bias (ml)
20.08
20.17
Limits of agreement (ml)
21.18–1.02
21.57–1.22
PEEP 6 cmH20 Bias (ml)
20.50
20.21
21.29
20.23
Limits of agreement (ml)
23.35–2.34
22.34–1.92
23.86–1.27
21.49–1.02
PEEP 9 cmH20Bias (ml)
21.00
20.52
22.29
20.41
Limits of agreement (ml)
24.66–2.66
22.75–1.72
26.35–1.78
21.53–0.07
PEEP 12 cmH20 Bias (ml)
21.60
20.48
23.04
20.56
Limits of agreement (ml)
24.75–1.56
22.61–1.65
28.37–2.29
22.20–1.08
doi:10.1371/journal.pone.0021003.t002
Figure 4. Bland-Altman Plot for comparison between VTcalcand VTPNTduring pre-lavage measurements. Shown are Line of Identity
(grey line), Bias (blue line), and 95%-Limits of Agreement (dashed line). Calibration was performed at each PEEP level in all animals. Data points
indicate the means of measurements in individual piglets at different PEEP levels.
doi:10.1371/journal.pone.0021003.g004
Tidal Volume Monitoring by Impedance Measurement
PLoS ONE | www.plosone.org5June 2011 | Volume 6 | Issue 6 | e21003