Extrapolation from ten sections can make CT-based quantification of lung
aeration more practicable
-- Electronic supplementary material --
A. W. Reske1*, A. P. Reske1*, H. A. Gast2, M. Seiwerts2, A. Beda1, U. Gottschaldt3, C.
Josten4, D. Schreiter5, N. Heller6, H. Wrigge7, M. B. Amato8
1
Department of Anesthesiology and Intensive Care Medicine, University Hospital Carl
Gustav Carus, Dresden, Germany; Departments of 2Diagnostic and Interventional Radiology,
3
Anesthesiology and Intensive Care Medicine, and 4Trauma and Reconstructive Surgery,
University Hospital Leipzig, Leipzig, Germany; 5Department of Surgery, Surgical Intensive
Care Unit, University Hospital Carl Gustav Carus, Dresden, Germany; 6Institute of
Informatics, University of Leipzig, Leipzig, Germany; 7Department of Anesthesiology and
Intensive Care Medicine, University of Bonn, Germany; 8Cardio-Pulmonary Department,
Pulmonary Divison, Hospital das Clinicas, University of Sao Paulo, Sao Paulo, Brazil.
*These authors contributed equally to this work
Correspondence should be addressed to:
Dr. med. Andreas Reske
Klinik und Poliklinik fuer Anaesthesiologie und Intensivtherapie
Universitaetsklinikum Carl Gustav Carus
Fetscherstrasse 74, D-01307 Dresden, Germany
e-mail: andreas.reske@uniklinikum-dresden.de
Phone: +49 176 14455000
Fax: +49 351 458 4336
Patients and Methods
We retrospectively analyzed complete lung CTs performed in emergency trauma patients as
part of the diagnostic work-up or in patients with respiratory problems treated in the intensive
care unit (ICU). Each CT had been requested by the treating physicians. All already
segmented CT data in our database were eligible for this study. The retrospective,
observational analysis of CT data was approved by the institutional Ethics Committee, which
waived the need for informed consent. We intentionally report only basic demographic data
because this work focuses on the extrapolation method.
CT scanning
Depending on availability, two multislice CT-scanners were used, either a Somatom Volume
Zoom (Siemens, Erlangen, Germany: 120 kV tube voltage, 165 mAs tube current, 4 x 2.5 mm
collimation), or a Philips MX8000 IDT 16 (Philips Medical Systems, Hamburg, Germany:
120 kV voltage, 170 mAs current, 16 x 1.5 mm collimation). Two different reconstruction
methods were used during routine clinical imaging. Contiguous images were reconstructed
with either 10 mm slice thickness and the enhancing filter “B60f” on the Siemens, or with 5
mm thickness and the standard filter “B” on the Philips scanner. All patients received contrast
material (120 ml Iopamidol, Schering, Berlin, Germany). In this retrospective study, the
degree of inspiration during CT could not be controlled: Spontaneously breathing patients
were asked to hold their breath after inspiration (without checking for compliance) during CT.
Mechanically ventilated patients were scanned during uninterrupted ventilation as is current
clinical practice. Calibration of the CT-scanners was performed using air and the
manufacturer’s standard phantom.
CT analysis
Because CT-images with gross motion artifacts are a potential problem for CT densitometry,
such data sets were not eligible. The distribution of pulmonary opacifications was classified
using all CT-sections according to Puybasset by two independent observers [ESM1]. The
Osiris software (University Hospital Geneva, Switzerland) was used for manual segmentation
of the lung parenchyma. The total lung volume (Vtotal) and mass (Mtotal), the pulmonary gas
volume (Vgas), and the volumes and masses of four differently aerated lung compartments
were calculated voxel-by-voxel according to established methods [ESM1-9]. The following
HU-ranges were used to define differently aerated lung compartments: nonaerated, -100 to
+100 HU; poorly aerated, -101 to -500 HU; normally aerated, -501 to -900 HU; and
hyperaerated, -901 to -1.000 HU. All subvolumes and –masses, as well as Vgas were
calculated as percentage of Vtotal or Mtotal, respectively.
Extrapolation
The extrapolation method uses ten CT-sections instead of all sections covering the whole lung
(all CT-sections). The assumption of linear lung boundaries is made and the lung volume
between two CT-sections treated as approximately conic section. In line with the only
previous study using this approach [ESM10], the most cranial and caudal CT-sections and a
further eight evenly spaced CT-sections between them were analyzed. In case it was
impossible to retrospectively select eight CT-sections at perfectly equal intervals between the
most cranial and caudal CT-sections, intervals were adapted to obtain an almost even
distribution. These ten CT-sections were analyzed using the standard segmentation and
densitometry methods and the results were extrapolated to the entire lung as illustrated in
Figure 1 of the main manuscript [ESM10, ESM11].
In order to study whether the number of CT-sections used for extrapolation could be reduced
even further, the procedure was repeated with five and three, respectively, evenly spaced CTsections in 45 patients.
To test whether the extrapolation method was equally accurate for assessing intra-individual
changes of volume and mass parameters between repeat CTs, we analyzed a subgroup of
patients for whom data from repeat CT were available.
Statistics
Data are given as median and range (minimum and maximum values) or, if distribution of
data was normal as confirmed by D'Agostino & Pearson omnibus and Shapiro-Wilk normality
tests, as mean and standard deviation (±SD). The agreement between extrapolated values and
the corresponding values calculated from all CT-sections was assessed according to Bland
and Altman [ESM12, ESM13]. Bias values were compared between subgroups with MannWhitney or Kruskal-Wallis tests. Bonferroni-corrected Mann-Whitney tests were used for
post-hoc testing. Statistical analysis was performed using SPSS 12.0 (SPSS Inc., Chicago, IL,
USA) and GraphPad Prism 5.0 (GraphPad Software, San Diego, California, USA). Statistical
significance was assumed if P < 0.05.
Results
We identified 173 eligible CT data sets in our data base. Some data sets had to be excluded
because of image reconstruction involving overlap (n=12) or non-standard image
reconstruction (n=4). Finally, we analyzed single helical CT data sets of 157 patients. For 26
of these patients, additional data from repeat CT were available.
The mean lung height (cranio-caudal distance) was 226 ± 33 millimeters. The mean distance
between the positions of CT-sections selected for extrapolation was 26 ± 4 millimeters. The
mean number of CT-sections covering the lung was 25±3 for 10 mm, and 47±7 for 5 mm
slices thickness.
In a subgroup of patients, the accuracy of the extrapolation method was compared using ten,
five and three CT-sections, respectively. Fifteen of these 45 patients had normal lungs. The
distribution of opacifications was lobar in eight, diffuse in 10, and patchy in 12 patients. For
Vtotal and Mtotal, bias was significantly greater for extrapolation from three as compared with
five or ten CT-sections, respectively. In contrast, bias for Vtotal and Mtotal did not differ
significantly when five or ten CT-sections were used for extrapolation (both P values greater
than 0.1). While bias values for relative subvolumes and –masses of differently aerated lung
compartments did not change significantly with the number of CT-sections used (all P values
greater than 0.05), the limits of agreement were considerably broader (Table 3 of the main
manuscript).
In 26 patients, additional data from repeat CTs were analyzed. The median time between
successive CTs was 2 (1-14) days. Absolute changes between two CTs and the agreement
between extrapolated changes and the corresponding values from the whole lung are shown in
table 4 of the main manuscript.
References for the Electronic Supplementary Material
ESM1:
Puybasset L, Cluzel P, Gusman P, Grenier P, Preteux F, Rouby JJ (2000)
Regional distribution of gas and tissue in acute respiratory distress syndrome. I.
Consequences for lung morphology. CT Scan ARDS Study Group. Intensive Care Med
26:857–869.
ESM2:
Gattinoni L, Caironi P, Pelosi P, Goodman LR (2001) What has computed
tomography taught us about the acute respiratory distress syndrome? Am J Respir Crit Care
Med 164:1701–1711.
ESM3:
Patroniti N, Bellani G, Maggioni E, Manfio A, Marcora B, Pesenti A (2005)
Measurement of pulmonary edema in patients with acute respiratory distress syndrome. Crit
Care Med 33:2547-2554.
ESM4:
Rouby JJ, Puybasset L, Nieszkowska A, Lu Q (2003) Acute respiratory distress
syndrome: lessons from computed tomography of the whole lung. Crit Care Med
31(Suppl):S285-295.
ESM5:
Caironi P, Carlesso E, Gattinoni L (2006) Radiological imaging in acute lung
injury and acute respiratory distress syndrome. Semin Respir Crit Care Med 27:404-415.
ESM6:
Borges JB, Okamoto VN, Matos GF, Caramez MP, Arantes PR, Barros F,
Souza CE, Victorino JA, Kacmarek RM, Barbas CS, Carvalho CR, Amato MB (2006)
Reversibility of lung collapse and hypoxemia in early acute respiratory distress syndrome.
Am J Respir Crit Care Med 174:268-278.
ESM7:
Rylander C, Högman M, Perchiazzi G, Magnusson A, Hedenstierna G (2004)
Oleic acid lung injury: a morphometric analysis using computed tomography. Acta
Anaesthesiol Scand 48:1123-1129.
ESM8:
Wrigge H, Zinserling J, Neumann P, Muders T, Magnusson A, Putensen C,
Hedenstierna G (2005) Spontaneous breathing with airw y pressure release ventilation favors
ventilation in dependent lung regions and counters cyclic alveolar collapse in oleic-acidinduced lung injury: a randomized controlled computed tomography trial. Crit Care 9:R780789.
ESM9:
Simon BA (2000) Non-invasive imaging of regional lung function using x-ray
computed tomography. J Clin Monit 16:433–442.
ESM10:
Rylander C, Tylén U, Rossi-Norrlund R, Herrmann P, Quintel M, Bake B
(2005) Uneven distribution of ventilation in acute respiratory distress syndrome. Crit Care
9:R165-171.
ESM11:
Reske AW, Gast HA, Seiwerts M, Kahn T, Gottschaldt U, Schreiter D, Josten
C, Amato MBP, Reske AP (2009) Evaluation of a method to quicken the CT-based
quantification of lung aeration for the entire lung. Intensive Care Med 35(Suppl 1):P302.
ESM12:
Bland JM, Altman DG (1986) Statistical methods for assessing agreement
between two methods of clinical measurement. Lancet 1(8476):307–310.
ESM13:
Dewitte K, Fierens C, Stöckl D, Thienpont LM (2002) Application of the
Bland–Altman plot for interpretation of method-comparison studies: a critical investigation of
its practice. Clin Chem 48:799–801.