COMPUTERIZED ESTIMATION OF THE LUNG VOLUME REMOVED DURING

COMPUTERIZED ESTIMATION OF THE LUNG VOLUME REMOVED DURING

LUNG VOLUME REDUCTION SURGERY

Sebastien Gilbert1,2, MD Bin Zheng3, PhD

Joseph K. Leader3, PhD James D. Luketich1, MD Carl R. Fuhrman4, MD

Rodney J. Landreneau1, MD David Gur3, ScD

Frank C. Sciurba5, MD

1Department of Surgery, University of Pittsburgh, Pittsburgh, PA, 15213

2Veterans Affairs Pittsburgh Health System, Pittsburgh, PA 15240

3Department of Radiology, University of Pittsburgh, Pittsburgh, PA, 15213

4Department of Radiology, University of Pittsburgh, Pittsburgh, PA, 15213

5Department of Medicine, University of Pittsburgh, Pittsburgh, PA, 15213 Address correspondence and reprint requests to: J. Ken Leader, Ph.D. University of Pittsburgh Department of Radiology Imaging Research Division 300 Halket Street, Suite 4200 Pittsburgh, PA 15213 Phone: (412) 641-2572, Fax: (412) 641-2582 Email: [email protected] Running head: LVRS Lung Volume Removed

* Manuscript Title Page

LVRS Lung Volume Removed

page 1

COMPUTERIZED ESTIMATION OF THE LUNG VOLUME REMOVED DURING LUNG

VOLUME REDUCTION SURGERY

ABSTRACT

Rationale and Objective. This study was designed to develop an automated method for estimating lung

volume removed during lung volume reduction surgery (LVRS) using computed tomography (CT).

Materials and Methods. The CT examinations of six patients who underwent bilateral LVRS were

analyzed in this study. The resected lung tissue (right and left) was weighed during pathological

examination. An automated computer scheme was developed to estimate the lung volume removed

using the CT voxel values and lung specimen weight. The computed fraction of lung volume removed

was evaluated across a range of simulated surgical planes (i.e., other than parallel to the CT image

plane) and CT reconstruction kernels, and it was compared to the surgeons’ post-surgical estimates.

Results. The computed fraction of the lung volume removed during LVRS was linearly correlated with

the resected lung tissue weight (Pearson correlation = 0.697, p = 0.012). The computed fraction of lung

volume removed ranged from 12.9% to 51.7% of the total lung volume. The surgeons’ post-surgical

estimates of lung volume removed ranged from 30 to 33%. The percent difference between the

surgeons’ estimates and the computed lung volume removed as a percentage of the surgeons’ estimates

ranged from -72.3% to 57.0% with mean absolute difference of 29.7% (± 20.7).

Conclusion. The preliminary findings of this study suggest that the proposed quantitative model should

provide an objective measure of lung volume removed during LVRS that may be used to investigate the

relationship between lung volume removed and outcome.

Keywords: Lung volume reduction surgery, computed tomography, emphysema, CT densitomtery.

* Manuscript (excl. author details)


page 2

INTRODUCTION

In genetically susceptible individuals, exposure to tobacco smoke and air pollution may lead to

chronic obstructive pulmonary disease (COPD), which is the 12th leading cause of disability worldwide

[1-2]. Chronic inflammation caused by exposure to these irritants may ultimately destroy the lung

parenchyma (e.g., emphysema) resulting in loss of lung elastic recoil [3]. The concept of lung volume

reduction surgery (LVRS) as a treatment for emphysema was originated by Oto Brantigan decades ago

[4], and later revived by Cooper et al. [5] in the mid 1990’s. Although the benefits and mortality

associated with LVRS have been debated [6,7], the National Emphysema Treatment Trial (NETT) and

other randomized studies have demonstrated that LVRS is an effective treatment option for patients with

advanced emphysema, particularly those with predominantly upper lobe disease and poor exercise

capacity [8-12]. Video-assisted thoracoscopic (VATS) and median sternotomy are the most common

surgical approaches and are reported to provide equal benefit to the patient [13]. The excision of non-

function lung tissue is believed to improve lung function by relieving compression on normal underlying

lung tissue, and to improve lung elastic recoil, which permits outward forces to restore collapse

bronchioles [5,14-16].

Lung function (e.g., Pulmonary function test [PFT]) and radiological assessments (subjective and

quantitative computed tomography [CT] analysis) are considered effective tools to detect and assess the

severity of emphysema and identify LVRS candidates [17,18]. While PFTs provide global information

regarding lung function, CT and ventilation-perfusion imaging can be used to determine the regional

distribution of emphysematous changes, a finding that is essential to proper patient selection for LVRS

[19]. For example, documentation of upper-lobe predominance is an essential selection criterion for

LVRS [11,20]. Using the chest X-ray, CT scan, and ventilation-perfusion scan the surgeon can

integrate the above information to form a mental picture of the plane along which the right and left lung

parenchyma will be resected.


page 3

A number of measurement tools, including morphological features extracted from CT images,

have been investigated to predict LVRS outcome [21-24], but the optimal predictor(s) of post-LVRS

outcome has yet to be identified in patients who meet current anatomic and functional criteria for LVRS.

The prognostic implications of excised lung volume on patient outcomes need to be further

characterized. An initial step in this direction is the development of a robust image analysis

methodology that can accurately estimate the fraction of lung volume removed. Although investigators

have suggested that the weight of resected lung tissue may be a determining factor in LVRS outcome

[25-28], it is not a practical parameter that the surgeon can integrate in the operative strategy compared

to the corresponding lung volume. The use of standardized CT density measurements provides a

relationship between weight and volume that may be clinically useful in prospectively examining their

potential effect on outcome. This study presents a computerized method that utilizes the preoperative

CT examination and the weight of the resected lung specimen to unilaterally estimate the lung volume

removed following LVRS.

MATERIALS AND METHODS

A. Subjects

The preoperative CT examinations of six consecutive emphysema patients who underwent

LVRS at our institution over a twelve month period were analyzed. All subjects were evaluated by a

multidisciplinary committee composed of a thoracic surgeon, pulmonologist, and thoracic radiologist.

Patients were selected for LVRS according to specific guidelines established by Medicare. The major

inclusion criteria include: (1) upper-lobe dominant emphysema on CT examination, (2) FEV1 at least

20% of predicted, (3) DLCO at least 20% of predicted. Preoperative data and six-month postoperative

data were prospectively collected on all subjects (Table 1). Pre- and post-operative lung volumes were

measured using body plethysmography.


page 4

Data collection for this study was performed under University of Pittsburgh Internal Review

Board (IRB) approved protocols. All examinations and image data were de-identified through removal

of Protected Health Information (PHI). A separate file containing the PHI, which was used to generate

summary tables, was maintained by a certified “Honest Broker” compliant with Health Insurance

Portability and Accountability Act (HIPAA) regulations and shielded from the investigators.

B. LVRS procedure

The LVRS procedure was performed through bilateral anterior submammary thoracotomy

incisions (n = 5) or via a bilateral 3-port thoracoscopic approach (n = 1). After reviewing the

preoperative imaging, the surgeon identified the target pulmonary parenchyma for resection. The

diseased lung was resected using mechanical staplers buttressed with resorbable strips (W.L. Gore &

Associates, Flagstaff, AZ). Once the specimen is removed from the pleural cavity, the weight is

measured and documented on the pathology report. The number of stapler cartridges used to resect each

specimen was noted in order to allow subtraction of the weight of the staples and buttressing strips from

the specimen weight. The weight of one buttressing strip was obtained by firing a surgical stapler

loaded with strips. The average weight of one buttressing strip across 10 measurements of the strip with

its rows of staples was 0.17g. The final weight of the lung tissue resected was calculated by subtracting

the weight of the buttressing strips from the pathologic specimen weight. After the LVRS procedure,

the surgeons was asked to estimate the volume of the right and left lungs removed as a percentage of

total lung volume.

C. CT examinations

The helical CT examinations were performed using a LightSpeed Plus scanner (GE Healthcare,

Waukesha, WI) at end inspiration. The acquisition parameters ranged from 120 to 140 KVP and 200 to

300 mAs. The CT images were contiguous volume scans reconstructed at a section thickness of 5.0 mm

with 512 × 512 pixel matrices using the high spatial frequency “lung” convolution kernel (GE


page 5

Healthcare). Computed tomography images reconstructed with additional kernels (i.e., “standard” and

“soft”) were used in a secondary analysis to measure the effect of reconstruction kernels. The pixel sizes

of the reconstructed CT images ranged from 0.51 mm to 0.89 mm.

D. Computation of removed lung volume

A multi-stage computerized scheme was developed to estimate the fraction of the lung volume

removed during an LVRS procedure using CT image data and the weight of the resected lung tissue in

the following steps: (1) segment the lungs depicted on CT images, (2) estimate the lung volume and

weight from CT voxel values of the segmented lungs, and (3) implement a set of simulated resection

surfaces to estimate the fraction (percent) of the lung volume removed by matching the computed

weight to the weight of the resected tissue.

Our automated lung segmentation computer scheme involves multiple steps of adaptive

thresholding, morphological filtering, and region labeling to identify CT pixels depicting lung and non-

lung tissue [29]. The unique characteristics of the simple and robust segmentation algorithm include: (1)

an adaptive threshold level computed for each image, which may be important in abnormal

examinations because the pixel value histograms could vary substantially from those in negative

examinations and (2) implementation of simple knowledge-based classification rules. The automated

segmentation scheme was evaluated under different CT scanning protocols (e.g., radiation exposure and

CT section thickness) [29,30]. Additionally, the total lung volume computed from CT examinations

using our automated segmentation method was correlated to body plethysmography.

After segmenting the lungs, the computer scheme estimates the lung volume and weight as

depicted in the CT image data. Lung volume is computed by multiplying the sum of the segmented lung

voxels by the voxel dimensions for each CT image. Total lung volume is the sum of the lung volume for

all CT image sections. Voxel density (VD) in the lung (g/ml) was computed directly from the pixel


page 6

values (HU) under the assumption that there is a linear relation between CT pixel value and tissue

density in the range of interest as:

.1024

1024+= HUVD

Estimated total lung weight is then computed as the sum of all voxel weights for all segmented lung

voxels. The computer scheme estimates unilateral lung weight image-by-image and records the

unilateral lung volume, average voxel density, cumulative (integrated) lung volume and weight from the

apex to the base of the lung.

To estimate the fraction (percent) of the lung volume removed during the LVRS procedure, the

weight of the resected tissue is compared to the computed cumulative lung volume that is assumed to

represent the resected tissue using a linear interpolation method. The simulated resection surface in this

analysis is assumed to be a plane (Fig. 1). The computed lung volume removed is defined as the

cumulative lung volume at the level when the computed cumulative lung weight equals the resected lung

tissue weight.

E. Error analysis

Two potential sources of errors using the proposed approach were investigated, which are the CT

reconstruction filter and the surgical approach (resection surface). There is no available tool to record

the exact three-dimensional resection surface created by the surgeon during an LVRS procedure and

relate the resection surface to CT image data. While some surgeons may perform the lung resection

close to the transverse plane of the lung, others may resect along a path parallel to the outside surface of

the lung (i.e. more curvilinear). To estimate the errors associated with the simulated surgical resection

surface, the simulated resection plane was varied relative to the CT image reconstruction plane. The

simulated resection planes for the left and right lungs were oriented oppositely (Fig. 1). Each simulated


page 7

surgical resection plane was moved caudally starting from the lung apex and cumulative lung volume

and weight were computed as described above.

The impact of selecting different CT image reconstruction kernels on the estimates was assessed

by comparing measurements using all the available reconstructions and filtering the original CT images.

In all six cases images were reconstructed using GE Healthcare’s “lung” (high-spatial frequency) kernel

and at least one additional kernel (i.e., “soft” or “standard”). A series of low-pass Gaussian filters were

applied to the CT images reconstructed with the “lung” kernel. The computer scheme was re-applied to

the CT examinations and the differences in computed lung volume removed were compared.

F. Data analysis

The computed lung volume and weight are reported unilaterally along with the computed

fraction of the lung volume removed during the LVRS procedure. The surgical resection surface was

simulated for 20 different planes from 0 to 38 degrees at 2 degree increments, and the mean and range of

computed lung volume removed was calculated (Fig. 1). The computed fraction of lung volume

removed is reported as the mean across the 20 measurements. Pearson correlation coefficients were

used to evaluate the relation between various parameters and tested for the difference from zero with p <

0.05 considered statistically significant. The difference between the postoperative estimates by the

surgeons and the computed lung volume removed were reported as a percentage of the surgeons’

estimates.

CT examinations from four subjects were reconstructed using three kernels (i.e., GE

Healthcare’s “standard”, “soft”, and “lung”), and the computed lung volume, lung weight, and lung

volume removed were calculated for each reconstruction. Additionally, these values were computed for

CT images reconstructed with the “lung” kernel that were filtered using a low-pass Gaussian filter with

kernel sizes of 5, 7, and 9 pixels.


page 8

RESULTS

Parameters representing lung characteristics computed from CT images were consistent with

pulmonary volume measurements and emphysema characteristics (i.e., upper-lobe predominance). Total

lung volume computed from CT images and total lung capacity (TLC) measured from body

plethysmography showed a trend towards linear correlation, but the correlation was not significantly

different from zero (Pearson correlation = 0.725, p = 0.104). There was a wide range of lung volume

and weight computed from the CT examinations across the twelve lungs of the six patients (Table 2)

similar to the range of TLC obtained from body plethysmography. The mean computed lung volume

removed was 27.7% (± 10.5).

Average computed density per CT image section increased from the apex to the base of the lung

(Fig. 2). However, there was a dramatic decrease in density for the initial image sections of the apices,

which may have resulted from partial volume averaging with the apical chest wall. The left lung in Case

3 did not present the same pattern. The computed density in this case only decreased slightly in the

middle of the lung.

The method for computing the lung volume removed during LVRS was consistent and robust for

a range of simulated surgical planes (Table 2) and reconstruction kernels (Table 3). The range of

computed lung volume removed across surgical resection planes (from 0 to 38 degrees) was small in all

cases. The minimum range and maximum range of computed lung volume removed across the six

subjects (twelve lungs) were 0.19% and 2.50%, respectively. Computed total lung weight was inversely

related to the sharpness (spatial frequency) of the reconstruction kernel, and, hence, the computed lung

volume removed was directly related to the sharpness of the reconstruction kernel (Table 3). The

disparity in computed lung volume removed between high- and low-spatial frequency kernels was

reduced by application of a low-pass Gaussian filter to CT images reconstructed using the high-spatial


page 9

frequency “lung” kernel (Table 3). The differences in computed lung volume removed across the

different reconstruction kernels and filtering were less than 5% for all four cases (data not shown).

The computed fraction of the lung volume removed during LVRS was and linearly correlated

with the resected lung tissue weight (Pearson correlation = 0.697, p = 0.012) (Fig. 3). The weight of

resected lung tissue ranged from 34 to 150 grams and the computed lung volume removed ranged from

12.9% to 51.7% for the 12 lungs of the six patients (Table 2). The variation between the computed lung

volume removed for the subject’s left and right lungs was consistent in four of the six cases (less than

8.3% difference). In two cases, the intra-case variations were 17.9% and 19.9%.

The percent difference between the surgeons’ postoperative estimates of lung volume removed

and the computed lung volume removed varied greatly. The surgeons estimated that approximately 33%

of lung volume was bilaterally removed from Cases 1 and 6 and that 30% was bilaterally removed from

Cases 2, 3, 4, and 5. The percent difference between the surgeons’ estimates and the computed lung

volume removed as a percentage of the surgeons’ estimates ranged from -72.3% to 57.0% with mean

absolute difference of 29.7% (± 20.7). The difference between the surgeons’ estimate and computed

lung volume removed was not correlated to the lung volume computed from CT (Pearson correlation =

0.067, p = 0.838).

DISCUSSION

The recommended extent of pulmonary resection during LVRS appears to be based largely on

surgical experience and dogma rather than empiric evidence of optimal effectiveness. This issue

deserves more rigorous and extensive investigation. The methodology described in this study provides

an objective measure to evaluate the impact of LVRS on emphysema patients and an additional level of

refinement beyond lung tissue weight by quantifying the volume of lung removed. The methodology

accounts for variability in tissue density between patients and within the lung parenchyma of each

individual patient.


page 10

Resected tissue weight has been investigated as an LVRS outcome measures in human and

animal models, but the weight of the resected tissue can be affected by several other factors (e.g., tissue

edema, fibrosis, nodules) [27,31]. In a rabbit model, an improvement in diffusion capacity measured

with carbon monoxide (DLCO) post-LVRS was observed up to a certain weight of resected tissue after

which DLCO decreased, which may indicate that DLCO is an appropriate outcome metric [25].

Expiratory flows, residual volumes, and lung elastic recoil have been identified as preoperative predictor

[21], but these metrics have also been questioned because of their linear relation to the weight of the

lung resected [25-27,32]. Mortality was not observed to correlate to lung tissue weight resected [28].

Since specimen weight alone does not account for the variability in disease severity and presentation, it

may be an imprecise surrogate for the actual lung volume removed during LVRS.

The relation, if any, between the fraction (percentage) of the lung volume removed during LVRS

and post-surgical outcome is not well understood. The accuracy of comparing pre- and postoperative

imaging to obtain such measurements is limited by several factors, including the degree of expansion of

the remaining lung, the reshaping of the diaphragm, and changes in chest wall configuration. Therefore,

postoperative CT scanning cannot reliably estimate the amount of the lung volume removed. Although

body plethysmography can measure volume changes, the test has limited usefulness because it requires a

fit and cooperative post-surgical patient. This preliminary study suggests that surgeon’s subjective

estimate of the fraction of the lung volume removed may not be reliable when compared to the

computerized CT analysis.

The need for objective measurement tools was also underscored by the lack of correlation

between the surgeons’ estimate of the lung volume removed and the computed value. The large

variability between the surgeon’s estimate and the computed value for individual cases resulted in a

significant difference between the two estimates. Holbert et al. [33] reported a similar large variability

between the percent of lung volume targeted for resection and the computed lung volume from pre- and


page 11

post-operative CT examinations. However, comparing pre- and postoperative CT images may not be as

accurate as our proposed model because of thoracic cavity changes described above.

In the present study, differences between the simulated and actual surgical resection surface may

be a factor contributing to the discrepancy observed between surgical estimates and computed volumes.

However, the wide range of simulated resection planes analyzed demonstrated that modifying the

simulated resection plane had a relatively small affect on the computed lung volume removed (Table 2).

Simulating the actual resection surface with a curvilinear surface may improve the model accuracy, but

this added level of refinement is not anticipated to make a significant difference.

There are additional limitations to our study results. First, the visceral pleura and, if present,

certain pleural lesions (e.g., pleural plaques) depicted on CT examination are typically not included as

part of the lung parenchyma segmented from the CT images. The visceral pleura and potentially

attached lesions removed during LVRS would be weighed as part of the surgical specimen and,

consequently, confound computation of the lung volume removed. This preliminary study did not

address these issues, but the lung segmentation could be extended by one or two pixels to compensate

for the visceral pleura and manual correction for resected visceral pleural lesions could be incorporated

into the methodology. Second, intra-pulmonary abnormalities (e.g., edema, consolidation, scars,

nodules) near the pleura also may not be segmented from the CT images as part of the lung parenchyma,

but could also be resected and weighed as part of the surgical specimen. Manual correction could also

be implemented to adjust for this potential problem. Third, CT reconstruction with a high-spatial

frequency kernel (i.e., GE Healthcare “lung”) may not be optimal for this quantitative analysis because

of the possibility for non-linear relationship between pixel values and density, but this was the only

reconstruction kernel common to all six cases. This systematic bias could be accounted for, at least

partially, by application of a low-pass filter to CT images reconstructed with high-spatial frequency

kernels prior to the application of the scheme, which reduced the bias to values comparable to low-


page 12

spatial frequency reconstruction kernels (Table 3). Finally, this preliminary study involved a small

number of subjects and, therefore, further investigation is necessary to establish correlation with lung

function and the robustness of the computer method to estimate the lung volume removed during LVRS.

In conclusion, the approach described in this study should provide an objective measure of lung

volume removed during LVRS that may be more accurate and consistent than the surgeon’s post-

surgical estimation. It should also be superior to image analysis methods that rely on pre- and

postoperative CT images comparisons to compute lung volume removed. A consistent and accurate

estimate of lung volume removed during LVRS may provide a valuable tool to facilitate refinements to

the planning and technical aspects of LVRS, and may provide additional data to investigate the relation

between LVRS technique and patient outcome.


page 13

REFERENCES

1. Murray CJL, Lopez AD. Evidence-based health policy--lessons from the Global Burden of Disease

Study. Science 274:740-743, 1996.

2. Pauwels RA, Buist AS, Calverley PM. Global strategy for the diagnosis, management, and

prevention of chronic obstructive pulmonary disease: NHLBI/WHO Global Initiative for chronic

obstructive lung disease (GOLD) workshop summary. Am J Respir Crit Care Med 163:1256-1276,

2001.

3. ATS Statement. Standards for the diagnosis and care of patients with chronic obstructive pulmonary

disease. Am J Respir Crit Care Med 152:S77-S83, 1995.

4. Brantigan OC. Mueller E. Surgical treatment of pulmonary emphysema. Am Surg 1957; 23:789-804.

5. Cooper JD, Trulock EP, Triantafillou AN, et al. Bilateral pneumectomy (volume reduction) for

chronic obstructive pulmonary disease. J Thorac Cardiovasc Surg 1995; 109:106-116.

6. Geddes D, Davies M, Koyama H, et al. Effect of lung-volume-reduction surgery in patients with

severe emphysema. N Engl J Med 2000; 343:239-245.

7. Hillerdal G, Lofdahl CG, Strom K, Skoogh BE, Jorfeldt L. Comparison of lung volume reduction

surgery and physical training on health status and physiologic outcomes: a randomized controlled

clinical trial. Chest 2005; 128:3489-3499.

8. Keenan RJ, Landreneau RJ, Sciurba FC, et al. Unilateral thoracoscopic surgical approach for diffuse

emphysema. J Thorac Cardiovasc Surg 1996; 111:308-315.

9. Yusen RD, Trulock EP, Pohl MS, Biggar DG. Results of lung volume reduction surgery in patients

with emphysema. The Washington University Emphysema Surgery Group. Semin Thorac

Cardiovasc Surg 1996; 8:99-109.

10. Naunheim KS, Hazelrigg SR, Kaiser LR, et al. Risk analysis for thoracoscopic lung volume

reduction: a multi-institutional experience. Eur J Cardiothorac Surg 2000; 17:673-679.


page 14

11. National Emphysema Treatment Trial Research Group. A randomized trial comparing lung-volume-

reduction surgery with medical therapy for severe emphysema. N Engl J Med 2003; 348:2059-2073.

12. Miller JD, Malthaner RA, Goldsmith CH, et al. A randomized clinical trial of lung volume reduction

surgery versus best medical care for patients with advanced emphysema: a two-year study from

Canada. Ann Thorac Surg 2006; 81:314-320.

13. National Emphysema Treatment Trial Research Group. Safety and efficacy of median sternotomy

versus video-assisted thoracic surgery for lung volume reduction surgery. J Thorac Cardiovasc Surg

2004 127:1350-1360.

14. Sciurba FC, Rogers RM, Keenan RJ, et al. Improvement in pulmonary function and elastic recoil

after lung-reduction surgery for diffuse emphysema. N Engl J Med 1996; 334:1095-1099.

15. Becker MD, Berkmen YM, Austin JH, et al. Lung volumes before and after lung volume reduction

surgery: quantitative CT analysis. Am J Respir Crit Care Med 1998; 157:1593-159.

16. Gelb AF, McKenna RJ Jr, Brenner M, Schein MJ, Zamel N, Fischel R. Lung function 4 years after

lung volume reduction surgery for emphysema. Chest 1999; 116:1608-1615.

17. Goldin JG. Quantitative CT of emphysema and the airways. J Thorac Imaging 2004; 19:235-240.

18. Zaporozhan J, Ley S, Eberhardt R, et al. Paired inspiratory/expiratory volumetric thin-slice CT scan

for emphysema analysis: comparison of different quantitative evaluations and pulmonary function

test. Chest 2005; 128:3212-3220.

19. Aziz ZA, Wells AU, Desai SR, et al. Functional impairment in emphysema: contribution of airway

abnormalities and distribution of parenchymal disease. AJR Am J Roentgenol 2005; 185:1509-1515.

20. Stavngarad T, Shaker SB, Dirksen A. Quantitative assessment of emphysema distribution in smokers

and patients with alpha (1)-antitrypsin deficiency, Respir Med 2006; 100:94-100.

21. Sciurba FC. Preoperative predictors of outcome following lung volume reduction surgery. Thorax

2002; 57(Suppl 2):ii47-ii52.


page 15

22. Coxson HO, Rogers RM, Whittall KP, et al. A quantification of lung surface area in emphysema

using computed tomography. Am J Respir Crit Care Med 1999; 159:851-856.

23. Leader JK, Roger RM, Fuhrman CR, et al. Size and morphology of the trachea before and after lung

volume reduction surgery. AJR Am J Roentgenol 2004; 183:315-321.

24. Rogers RM, Coxson HO, Sciurba FC, Keenan RJ, Whittall KP, Hogg JC. Preoperative severity of

emphysema predictive of improvement after lung volume reduction surgery: use of CT

morphometry. Chest 2000; 118:1240-1247.

25. Chen JC, Brenner M, Huh J, et al. Effect of lung volume reduction surgery on pulmonary diffusion

capacity in a rabbit model of emphysema. J Surg Res 1998; 78:155-160.

26. Huh J, Brenner M, Chen JC, et al. Changes in pulmonary physiology after lung volume reduction

surgery in a rabbit model of emphysema. J Thorac Cardiovasc Surg 1998; 115:328-334.

27. Brenner M, McKenna RJ Jr, Chen JC, et al. Relationship between amount of lung resected and

outcome after lung volume reduction surgery. Ann Thorac Surg 2000; 69:388-393.

28. Chen JC, Powell LL, Serna DL, et al. Pulmonary artery pressure: an intraoperative guide to limiting

resection volume. J Surg Res 1999; 82:137-145.

29. Leader JK, Zheng B, Rogers RM, et al. Automated lung segmentation in X-ray computed tomography:

Development and evaluation of a heuristic threshold-based scheme. Acad Radiol 2003; 10:1224-1236.

30. Zheng B, Leader JK, Fuhrman CR, Sciurba FC, Gur D. Automated detection and classification of

interstitial lung diseases from low dose CT images. Proc SPIE 2004; 5370:849-856.

31. Keller CA, Naunheim KS, Osterloh J, Espiritu J, McDonald JW, Ramos RR. Histopathologic

diagnosis made in lung tissue resected from patients with severe emphysema undergoing lung

volume reduction surgery. Chest 1997; 111:941-947.

32. McKenna RJ Jr, Brenner M, Fischel RJ, Gelb AF. Should lung volume reduction for emphysema be

unilateral or bilateral?. J Thorac Cardiovasc Surg 1996; 112:1331-1338.


page 16

33. Holbert JM, Brown ML, Sciurba FC, Keenan RJ, Landreneau RJ, Holzer AD. Changes in lung

volume and volume of emphysema after unilateral lung reduction surgery: analysis with CT lung

densitometry. Radiology 1996; 201:793-797.


page 17

Table Legends

Table 1

Subject demographics and preoperative and 6 months postoperative data

Table 2

Resected lung tissue weight measured from pathology, lung volume and lung volume removed

computed from CT examination

Table 3

CT image variable computed for different GE Healthcare reconstruction kernels and the “lung” kernel

reconstructions filtered using a series of Gaussian filters for Case 2


page 18

Figure Legend

Figure 1

Illustration of the approximation of the first five simulated surgical resection planes on a 2-D lung

projection with 0 degrees overlaid on the right lung and 38 degrees on the left lung.

Figure 2

Mean lung density per CT image for the left (• ) and right lung ( ) of a 67-year-old man with upper-lobe

dominant emphysema (Case 2). The CT images represent the lung from apex (low image numbers) to

the base (high image numbers).

Figure 3

Resected lung tissue weight versus computed fraction of the lung volume removed with the linear

regression line (R2 = 0.485).

Table 1 Subject demographics and preoperative and 6 months postoperative data

TLC (L, % predicted) FEV1 (L, % predicted) DLCO (L, % predicted) case gender age pre-op post-op pre-op post-op pre-op post-op 1 M 61 8.9 (121) 8.0 (109) 1.42 (40) 1.71 (49) 15.8 (55) 16.9 (58) 2 M 67 6.9 (109) 6.1 (97) 0.77 (26) 0.94 (32) 8.5 (38) 8.3 (37) 3 F 69 6.6 (136) 4.8 (98) 0.72 (35) 1.28 (64) 7.0 (37) 8.0 (43) 4 F 66 5.7 (165) pending 0.35 (23) pending 3.9 (25) pending 5 F 67 6.0 (129) 5.5 (118) 0.57 (29) 0.77 (39) 5.0 (30) 6.1 (36) 6 F 52 4.3 (95) 3.9 (86) 0.66 (29) 0.87 (38) 6.8 (37) 8.6 (45) TLC, total lung capacity

FEV1, forced expiratory volume in one second

DLCO, diffusion capacity measured with carbon monoxide

Table

Table 2 Resected lung tissue weight measured from pathology, lung volume and lung volume removed computed from CT examination values computed from CT examinations resected lung total lung fraction of lung volume removed (%) case lung tissue weight (g) volume (L) meana 0° - 38° range 1 left 80.0 4.5 20.2 19.8 – 20.5 right 100.0 4.2 28.5 28.0 – 28.9 2 left 150.0 3.0 51.7 51.2 – 52.3 right 120.0 3.7 33.8 33.5 – 34.1 3 left 57.8 1.4 38.3 37.5 – 39.4 right 64.6 1.7 18.4 18.3 – 18.5 4 left 59.0 2.9 31.7 30.5 – 32.8 right 69.0 3.1 28.4 27.7 – 29.2 5 left 42.0 3.1 12.9 12.6 – 13.2 right 47.5 3.8 19.7 18.9 – 20.3 6 left 33.7 1.3 21.4 20.2 – 22.7 right 35.0 1.8 27.9 27.5 – 28.2 amean across the simulated surgical resection planes from 0° to 38°

Table

Table 3 CT image variable computed for different GE Healthcare reconstruction kernels and the “lung” kernel reconstructions filtered using a series of Gaussian filters for Case 2

left lung right lung

CT kernel volume (L) weight (g) fraction of lung removed (%) volume (L) weight (g)

fraction of lung removed (%)

soft 3.00 398.2 47.4 3.67 507.9 30.7 standard 3.02 390.9 48.9 3.68 496.7 31.5 lung 3.02 361.1 51.7 3.68 461.1 33.8 lung-5a 3.01 374.2 50.7 3.68 480.1 33.1 lung-7a 3.02 398.0 47.4 3.68 508.5 30.6 lung-9a 3.01 406.2 46.8 3.68 520.3 30.1 alung-5, 7, and 9 represent images reconstructed using a “lung” kernel filtered using a Gaussian filter of kernel sizes of 5, 7, and 9 pixels, respectively.

Table

FigureClick here to download high resolution image

http://ees.elsevier.com/arad/download.aspx?id=22095&guid=e6aaf240-59a4-47af-ab68-0633dfda9100&scheme=1


http://ees.elsevier.com/arad/download.aspx?id=22096&guid=bd4bbd18-527d-4c82-993e-39a73938ce17&scheme=1


http://ees.elsevier.com/arad/download.aspx?id=22097&guid=7e152214-d4c1-499e-a7bb-40a67e3b6649&scheme=1

COMPUTERIZED ESTIMATION OF THE LUNG VOLUME REMOVED DURING

Documents