HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2282020088v1 228/2/577    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chiang, E. E. Articles by Simon, M. Search for Related Content PubMed PubMed Citation Articles by Chiang, E. E. Articles by Simon, M.

Detection of Pulmonary Embolism: Comparison of Paddlewheel and Coronal CT Reformations—Initial Experience¹

¹ From the Department of Radiology, Beth Israel Deaconess Medical Center, 330 Brookline Ave, Boston, MA 02215. From the 2001 RSNA scientific assembly. Received February 12, 2002; revision requested April 17; final revision received October 28; accepted November 25. Address correspondence to P.M.B. (e-mail: pboisell@caregroup.harvard.edu).

	ABSTRACT

TOP ABSTRACT INTRODUCTION Materials and Methods Results Discussion REFERENCES

 
In five patients with acute multilobar pulmonary embolism (PE) who were imaged with multi–detector row CT angiography, maximum intensity projection images were reformatted from axial images into rotated paddlewheel and coronal planes with three slab thicknesses and were reviewed for evidence of PE on a per-vessel basis with consensus of two readers. Paddlewheel reformations had a significantly higher percentage of overall detection of PE than did coronal reformations obtained with equivalent slab thickness (P < .0001). Paddlewheel reformations with 5.0-mm slab thickness had no significantly different percentage of overall detection of PE compared with that of axial images obtained with 2.5-mm collimation.

© RSNA, 2003

Index terms: Computed tomography (CT), angiography, 60.12116 • Computed tomography (CT), maximum intensity projection, 60.12119 • Computed tomography (CT), multi–detector row, 60.12119 • Embolism, pulmonary, 60.72 • Pulmonary arteries, CT, 60.12116

	INTRODUCTION

TOP ABSTRACT INTRODUCTION Materials and Methods Results Discussion REFERENCES

 
Computed tomographic (CT) pulmonary angiography is gaining widespread acceptance as a first-line study for the diagnosis of pulmonary embolism (PE) (1). Multi–detector row helical CT is an important recent advance in helical CT technology (2–8). Compared with single–detector row CT, advantages of multi–detector row CT include faster scanning time, decreased respiratory and cardiac motion, improved vascular conspicuity, and enhanced quality of reformatted images (2–8). These advantages have especially important implications for CT angiography. A relative disadvantage of multi–detector row CT is that it generates large data sets. Recently, there has been considerable interest in the use of multiplanar volume reformation (MPVR) methods in CT angiography, both as a mechanism for enhancing display of vascular structures and for improving the efficiency of interpretation by decreasing the number of images necessary for review (9).

With regard to CT pulmonary angiography, a limitation of MPVR images is that the traditional coronal and sagittal reformation planes are not ideal for following the branching patterns of the pulmonary arteries. Most of the images tend to "slice" the pulmonary vessels into small fragments of varying lengths. Recently, a new image display method has been described in which a set of planar slabs is arranged in a paddlewheel or rotational pattern that pivots on a central horizontal axis between the lung hila (Fig 1a, 1b) (10). The paddlewheel reformation technique provides a continuous display of branching vessels that radiate from both hila (Fig 1c) (10). However, the optimal slab thickness for the paddlewheel method and its accuracy for detection of PE have not yet been established.

View larger version (86K): [in this window] [in a new window] [Download PPT slide]   Figure 1a. Paddlewheel reformation method. (a) Drawing shows how paddlewheel slabs pivot on a central horizontal axis between the lung hila. (Reprinted, with permission, from reference 10.) (b) Lateral scout image for paddlewheel reformations with central axis at main pulmonary artery bifurcation. (c) Paddlewheel image at level of main pulmonary artery bifurcation shows continuous display of branching vessels that radiate from both hila.

View larger version (166K): [in this window] [in a new window] [Download PPT slide]   Figure 1b. Paddlewheel reformation method. (a) Drawing shows how paddlewheel slabs pivot on a central horizontal axis between the lung hila. (Reprinted, with permission, from reference 10.) (b) Lateral scout image for paddlewheel reformations with central axis at main pulmonary artery bifurcation. (c) Paddlewheel image at level of main pulmonary artery bifurcation shows continuous display of branching vessels that radiate from both hila.

View larger version (135K): [in this window] [in a new window] [Download PPT slide]   Figure 1c. Paddlewheel reformation method. (a) Drawing shows how paddlewheel slabs pivot on a central horizontal axis between the lung hila. (Reprinted, with permission, from reference 10.) (b) Lateral scout image for paddlewheel reformations with central axis at main pulmonary artery bifurcation. (c) Paddlewheel image at level of main pulmonary artery bifurcation shows continuous display of branching vessels that radiate from both hila.

	Materials and Methods

TOP ABSTRACT INTRODUCTION Materials and Methods Results Discussion REFERENCES

 
Patients and Imaging
Our radiology information system was used to retrospectively identify five patients with acute multilobar PE who were imaged with multi–detector row CT pulmonary angiography. Four patients were women, and one was a man, and their ages ranged from 37 to 82 years (mean age, 69 years). All five patients had acute shortness of breath, and their CT studies were performed within a 6-month period between May 2000 and January 2001. For the purpose of this study, we chose patients who had no substantial underlying lung or pleural disease, and their CT findings showed good opacification of the pulmonary arteries with little to no motion artifact, thus allowing for high-quality reformations. There were five patients who met these criteria. We received approval from our institutional review board to review findings of the patients’ imaging studies. Informed consent was not required.

Axial images had been obtained with a four-detector multi–detector row helical CT scanner (LightSpeed QX/i; GE Medical Systems, Milwaukee, Wis) by using 2.5-mm collimation, 1.25-mm reconstruction interval, 7.5-mm table speed per rotation, and 0.8-second gantry rotation time after intravenous administration of 75 mL of contrast material (ioversol injection 68%, Optiray 320; Mallinckrodt, St Louis, Mo) at 3.5 mL/sec. A small bolus (15 mL) test injection was administered to determine the optimal imaging delay for maximal pulmonary artery enhancement, which ranged from 10 to 20 seconds. Patients were scanned in a caudal to cranial direction to minimize the effect of respiratory motion.

CT Reformations
All reformations were performed by a senior CT technologist (K.F.R.). From the axial CT data sets, MPVR maximum intensity projection images were reformatted in both rotational paddlewheel planes and coronal planes. For each patient, three sets of paddlewheel images and three sets of coronal images were generated with 15.0-, 7.5-, and 5.0-mm slab thicknesses (Table 1). MPVR average intensity projection images were also obtained initially, but our preliminary review found that these data sets provided poor visualization of the pulmonary arteries outside of the hila. Thus, we did not perform an analysis of these data sets in the study.

Statistical Analysis
Statistical analysis was performed by using the McNemar test (12). This test was used to compare the percentage of overall detection of PE achieved with paddlewheel and coronal reformations obtained with equivalent slab thicknesses and to compare the percentage of overall detection of PE achieved with the individual reformation methods performed with differing slab thicknesses. A difference with a P value of less than .05 was considered statistically significant. We used statistical software (SAS, version 8.0; SAS Institute, Cary, NC).

	Results

TOP ABSTRACT INTRODUCTION Materials and Methods Results Discussion REFERENCES

 
After integration of the data from review of all axial, paddlewheel, and coronal data sets obtained in the five patients, a total of 96 pulmonary emboli were detected, including 13 emboli at the central pulmonary artery level, 22 emboli at the lobar pulmonary artery level, and 61 emboli at the segmental pulmonary artery level. Six emboli were detected only on axial data sets, six emboli were detected only on paddlewheel data sets, and one embolus was detected only on coronal data sets.

As shown in Table 2, the percentage of overall detection of PE for both paddlewheel and coronal MPVR methods significantly improved with decreasing slab thickness. With the paddlewheel reformations, the percentage of overall detection of PE was 63% (60 of 96) with the 15.0-mm-thick slabs, which improved to 84% (81 of 96) with the 5.0-mm-thick slabs (P < .0001) (Figs 2, 3). With the coronal reformations, the percentage of overall detection of PE was 38% (36 of 96) with the 15.0-mm-thick slabs, which improved to 54% (52 of 96) with the 5.0-mm-thick slabs (P < .0001). All of these differences were statistically significant, except for the differences obtained when the coronal 7.5-mm-thick slabs were compared with the coronal 5.0-mm-thick slabs (P = .0625) (Table 2). The number of discordant pairs between different slab thicknesses for each reformation method that were used for statistical analysis are listed in Table 2.

View larger version (94K): [in this window] [in a new window] [Download PPT slide]   Figure 2a. Reformatted paddlewheel images. (a) Paddlewheel image obtained with 15.0-mm-thick slab shows no embolus. (b) Paddlewheel image obtained with 5.0-mm-thick slab shows well-visualized embolus (arrow) in the distal left pulmonary artery.

View larger version (90K): [in this window] [in a new window] [Download PPT slide]   Figure 2b. Reformatted paddlewheel images. (a) Paddlewheel image obtained with 15.0-mm-thick slab shows no embolus. (b) Paddlewheel image obtained with 5.0-mm-thick slab shows well-visualized embolus (arrow) in the distal left pulmonary artery.

View larger version (125K): [in this window] [in a new window] [Download PPT slide]   Figure 3a. Paddlewheel versus coronal and axial images show improved efficiency of review and enhanced assessment of embolic burden. (a) Paddlewheel image shows the full extent of embolism (arrow) in posterior basal segment and subsegmental branches. (b) Four coronal images are required for visualization of extent of embolism (arrow). These images result in decrease in connectivity of vessels compared with paddlewheel display. (c) Multiple axial images are required for visualization of extent of embolism (arrow). These images result in further decrease in connectivity of vessels.

View larger version (24K): [in this window] [in a new window] [Download PPT slide]   Figure 3b. Paddlewheel versus coronal and axial images show improved efficiency of review and enhanced assessment of embolic burden. (a) Paddlewheel image shows the full extent of embolism (arrow) in posterior basal segment and subsegmental branches. (b) Four coronal images are required for visualization of extent of embolism (arrow). These images result in decrease in connectivity of vessels compared with paddlewheel display. (c) Multiple axial images are required for visualization of extent of embolism (arrow). These images result in further decrease in connectivity of vessels.

View larger version (98K): [in this window] [in a new window] [Download PPT slide]   Figure 3c. Paddlewheel versus coronal and axial images show improved efficiency of review and enhanced assessment of embolic burden. (a) Paddlewheel image shows the full extent of embolism (arrow) in posterior basal segment and subsegmental branches. (b) Four coronal images are required for visualization of extent of embolism (arrow). These images result in decrease in connectivity of vessels compared with paddlewheel display. (c) Multiple axial images are required for visualization of extent of embolism (arrow). These images result in further decrease in connectivity of vessels.

	Discussion

TOP ABSTRACT INTRODUCTION Materials and Methods Results Discussion REFERENCES

Our second aim was to compare the percentage of overall detection of PE achieved with the paddlewheel reformation method with that achieved with the coronal reformation methods. We found that the paddlewheel method was significantly superior to the coronal method at equivalent slab thicknesses. On the basis of the continuous display of vessels afforded by the paddlewheel technique compared with the more fragmented display with coronal images, this finding is not surprising. At 5.0-mm slab thickness, the paddlewheel images had a percentage of overall detection of PE that was similar to that of conventional axial images.

With MPVR techniques, the number of images for review was reduced several-fold, compared with the number of images produced with axial CT. If one anticipates that the trend toward use of narrower collimation for CT angiography will continue, the difference in the number of images will become even more significant. For example, a reduction in collimation from 2.50 to 1.25 mm would effectively double the number of images in an axial CT data set, with the assumption that one chooses to maintain the same reconstruction interval. In a recent study, Schoepf et al (13) advocated the use of 1-mm collimation to enhance detection of subsegmental emboli. Furthermore, viewing of the data sets with multiple window settings would also increase the number of images. In our study, we used standard mediastinal window settings to review all images, but some investigators have suggested that additional use of alternative window settings may be advantageous. For example, Raptopoulos and Boiselle (8) showed that use of a modified window setting (window width, 1,000 HU; window level, -100 HU) significantly increased pulmonary arterial conspicuity with CT angiographic studies. In our study, an average of 163 axial chest images were obtained. If one were to review these images with mediastinal, lung, and modified window settings, the number of images would increase to 489 (3 x 163).

Although we recognize that axial CT angiography has a high sensitivity and specificity for detection of PE, it is not a perfect method. For example, vessels coursing oblique to the axial plane are more readily assessed with reformatted images than they are with axial images. Our methods allowed us to assess whether reformatted images might depict emboli additional to those depicted with axial images. In this regard, there were six additional emboli depicted with the paddlewheel method, and this observation suggests a possible complementary role of the axial images. Although the paddlewheel MPVR method did not result in improved accuracy compared with that of axial CT, it does offer potential advantages in its continuous display of branching vessels from the hila to the lung periphery and in terms of the reduction in the number of images. This should improve diagnostic efficiency. Technically, it is a simple protocol, and a trained CT technologist can perform these reformations at the CT console with existing software in only a few minutes.

The current findings are promising, but our study has limitations. First, the study lacks a reference standard, such as conventional pulmonary angiography. Invasive pulmonary angiography is now only rarely performed at our institution and thus could not be used as a reference standard. Moreover, investigators in recent reports describe suboptimal interobserver agreement for this method (14,15), and this finding suggests that this test is not an ideal reference standard for detection of PE. In this study, we compared one data set against the other, rather than assessing the overall accuracy of CT for detection of PE. We thus produced our own aggregate standards against which each data set was compared. Second, the term "percent overall detection of PE," as used in our method, requires clarification. In the absence of a definitive standard for PE, we acknowledge that a true detection rate of PE cannot be determined. We are actually measuring the percentage of the aggregate of all detected PE that were found with each method. However, the differences in the percentage of overall detection of PE achieved with the various methods can be assessed by means of the McNemar test, which we used for statistical analysis. We further acknowledge that the reviewers were not blinded to the association between patient and image. This conceivably resulted in a lack of independent observations, and this lack weakened the statistical analysis. Third, this study included a small number of patients (n = 5). However, the number of presumed PE (filling defects) was large (n = 96) and thus allowed our findings to have statistical significance. We chose patients with multilobar PE, because even in the absence of an independent reference standard, there was no doubt about the presence or absence of emboli in this population. Our selection criteria resulted in an idealized subset of patients in whom image quality was excellent, and there were no substantial coexisting lung or pleural abnormalities. Because we were comparing different methods of viewing image data sets, it was important for the data to be free of artifacts. Finally, this study did not assess subsegmental emboli. Because the significance of isolated subsegmental emboli is uncertain, we decided to focus on areas of clinically relevant emboli. Although we did not formally assess subsegmental emboli, we noticed that subsegmental vessels were readily visualized with the paddlewheel technique. We suspect that this method will be helpful in assessment of emboli at this level. Future studies are needed for assessment of the paddlewheel method for detection of PE in a larger number of patients who have lesser embolic burdens and also for assessment of subsegmental emboli.

Although future studies are necessary before one can consider replacement of axial images with MPVR methods, there are sufficient data at present to suggest a complementary role. In an earlier study in which the role of reformations in the diagnosis of PE was assessed, Remy-Jardin et al (16) showed that oblique two-dimensional reformations enabled confident exclusion of PE at inconclusive CT examinations and improved the evaluation of the extent of thromboembolic disease. Similarly, the paddlewheel method may potentially serve as a complement to standard axial CT. Because the paddlewheel images demonstrate vascular continuity to a greater degree than do images obtained with other methods, they have the potential to improve reader confidence because they can be used to confirm a suspected PE depicted on axial CT images. Reader confidence assessments should be incorporated into future studies of this technique. Moreover, because the paddlewheel images display the craniocaudad extent of emboli in an efficient and visually accessible manner, we suspect that this technique will aid assessment of embolic burden. For example, the full extent of an embolus can often be demonstrated on a single paddlewheel image, but often several coronal or multiple sequential axial images are required to demonstrate the same entity (Fig 3). At our institution, we routinely perform paddlewheel reformations with all multi–detector row CT pulmonary angiographic studies, and we review them in conjunction with the axial images.

In summary, the percentage of overall detection of PE with MPVR images improves with decreasing slab thickness. For detection of PE, paddlewheel reformations are superior to coronal reformations, and paddlewheel reformations with 5.0-mm slab thickness are comparable to axial CT images. However, the paddlewheel method requires fewer images and provides a continuous rather than a fragmented display of vessels, and this could prove to be a diagnostic advantage in future studies.

	ACKNOWLEDGMENTS

 
We thank Roger B. Davis, ScD, for his assistance with statistical analysis.

	FOOTNOTES

 
Abbreviations: MPVR = multiplanar volume reformation, PE = pulmonary embolism

Author contributions: Guarantors of integrity of entire study, E.E.C., P.M.B.; study concepts and design, E.E.C., V.R., P.M.B.; literature research, E.E.C., P.M.B.; clinical studies, E.E.C., K.F.R., P.M.B.; data acquisition, K.F.R.; data analysis/interpretation, E.E.C., M.P.R., P.M.B.; statistical analysis, M.P.R.; manuscript preparation and definition of intellectual content, E.E.C., P.M.B.; manuscript editing, E.E.C., V.R., M.P.R., M.S., P.M.B.; manuscript revision/review and final version approval, all authors.

	REFERENCES

TOP ABSTRACT INTRODUCTION Materials and Methods Results Discussion REFERENCES

 

1. Goodman LR, Lipchik RJ, Kuzo RS, Liu Y, McAuliffe TL, O’Brien DJ. Subsequent pulmonary embolism: risk after a negative helical CT pulmonary angiogram—prospective comparison with scintigraphy. Radiology 2000; 215:535-542.[Abstract/Free Full Text]
2. Hu H, He HD, Foley WD, et al. Four multidetector-row helical CT: image quality and volume coverage speed. Radiology 2000; 215:55-62.[Abstract/Free Full Text]
3. Kopecky KK, Buckwalter KA, Sokiranski R. Multi-slice CT spirals past single-slice CT in diagnostic efficacy. Diagn Imaging 1999; 4:36-42.
4. Klingenbeck-Regn K, Schaller S, Flohr T, et al. Subsecond multi-slice computed tomography: basics and applications. Eur J Radiol 1999; 31:110-124.[CrossRef][Medline]
5. Rigauts H. A one year experience with the multislice helical CT. JBR-BTR 1998; 82:273-276.
6. Rubin GD, Shiau MC, Schmidt AJ, et al. Computed tomographic angiography: historical perspective and new state-of-the-art using multi detector-row helical computed tomography. J Comput Assist Tomogr 1999; 23(suppl 1):S83-S90.
7. Rubin GD, Shiau MC, Leung AN, et al. Aorta and iliac arteries: single versus multiple detector-row helical CT angiography. Radiology 2000; 215:670-676.[Abstract/Free Full Text]
8. Raptopoulos V, Boiselle PM. Multi-detector row spiral CT pulmonary angiography: comparison with single-detector row spiral CT. Radiology 2001; 221:606-613.[Abstract/Free Full Text]
9. Rubin GD. Data explosion: the challenge of multidetector-row helical CT. Eur J Radiol 2000; 36:74-80.[CrossRef][Medline]
10. Simon M, Boiselle PM, Choi JR, Rosen MP, Reynolds K, Raptopoulos V. Paddle-wheel CT display of pulmonary arteries and other lung structures: a new imaging approach. AJR Am J Roentgenol 2001; 177:195-198.[Free Full Text]
11. Remy-Jardin M, Remy J. Spiral CT angiography of the pulmonary circulation. Radiology 1999; 212:615-636.[Abstract/Free Full Text]
12. Dawson-Saunders B, Trapp RG. Basic and clinical biostatistics 2nd ed. East Norwalk, Conn: Appleton & Lange, 1994; 155-156.
13. Schoepf UJ, Holzknecht N, Helmberger TK, et al. Subsegmental pulmonary emboli: improved detection with thin-collimation multi-detector row spiral CT. Radiology 2002; 222:483-490.[Abstract/Free Full Text]
14. Diffin D, Leyendecker JR, Johnson SP, Zucker RJ, Grebe PJ. Effect of anatomic distribution of pulmonary emboli on interobserver agreement in the interpretation of pulmonary angiography. AJR Am J Roentgenol 1998; 171:1085-1089.[Abstract/Free Full Text]
15. Stein PD, Henry JW, Gottschalk A. Reassessment of pulmonary angiography for the diagnosis of pulmonary embolism: relation of interpreter agreement to the order of the involved pulmonary arterial branch. Radiology 1999; 210:689-691.[Abstract/Free Full Text]
16. Remy-Jardin M, Remy J, Cauvain O, Petyt L, Wannebroucq J, Beregi J. Diagnosis of central pulmonary embolism with helical CT: role of two-dimensional multiplanar reformations. AJR Am J Roentgenol 1995; 165:1131-1138.[Abstract/Free Full Text]

This article has been cited by other articles:

				S. Patel and E. A. Kazerooni Helical CT for the Evaluation of Acute Pulmonary Embolism Am. J. Roentgenol., July 1, 2005; 185(1): 135 - 149. [Abstract] [Full Text] [PDF]

				J H Reid Multislice CT pulmonary angiography and CT venography Br. J. Radiol., December 1, 2004; 77(suppl_1): S39 - S45. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2282020088v1 228/2/577    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chiang, E. E. Articles by Simon, M. Search for Related Content PubMed PubMed Citation Articles by Chiang, E. E. Articles by Simon, M.