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) Alert me when this article is cited 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 Garg, K. Articles by Barón, A. E. Search for Related Content PubMed PubMed Citation Articles by Garg, K. Articles by Barón, A. E. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?

Thromboembolic Disease

Variability of Interobserver Agreement in the Interpretation of CT Venography with CT Pulmonary Angiography

¹ Department of Radiology, Veterans Affairs Medical Center and University of Colorado, 1055 Clermont St., Denver, CO 80220.
² Present address: Diversified Radiology of Colorado, P. C., 1601 E. 19th Ave., Denver, CO 80218.
³ Department of Preventive Medicine and Biometrics, School of Medicine, University of Colorado Health Sciences Center, 4200 E. 9th Ave., Denver, CO 80262.

Received July 18, 2000; accepted after revision September 29, 2000.

 
Partially supported by a grant from the Society for Thoracic Radiology Research and Education Fund.

Address correspondence to K. Garg.

Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
OBJECTIVE. The objective of this study was to determine interobserver agreement in the diagnosis of acute deep venous thrombosis on CT venography performed in addition to CT pulmonary angiography.

SUBJECTS AND METHODS. One hundred forty-six CT venograms of 144 patients (mean age, 61.74 years) clinically suspected of having pulmonary embolism were analyzed prospectively and independently by two experienced thoracic and body imaging radiologists and later by consensus of the two radiologists. The CT venography protocol consisted of 5-mm-thick axial images at 20-mm intervals from the popliteal fossa to the renal veins. Images were acquired 3-4 min after the start of 100-150 mL of undiluted contrast medium administration at 4 mL/sec. Thirteen venous segments were analyzed in each patient. There were 1586 analyzable venous segments.

RESULTS. Interobserver agreement, with the patient as the unit of analysis, was moderately good (, 0.59; 95% confidence interval [CI], 0.39-0.78). Kappa values were similar for CT venography studies performed with 150 mL of contrast medium and 4-min delay (, 0.62; 95% CI, 0.30-0.88) and with 3-min delay and 100 mL of contrast medium (, 0.56; 95% CI, 0.32-0.80). Interobserver disagreement occurred in 17 (12%) of 146 CT venography studies. Findings of 11 CT venography studies were interpreted as negative, and six were interpreted as positive after consensus interpretation.

CONCLUSION. Interobserver agreement for deep venous thrombosis with CT venography is moderately good.

Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Venous thromboembolism is a major national health problem, especially among elderly patients. The reported annual incidence of venous thromboembolism varies, ranging from 43.7 to 145.0 per 100,000 patients for deep venous thrombosis (DVT) and 20.8 to 65.8 per 100,000 patients for pulmonary embolism [1,2,3]. Pulmonary embolism, a potentially fatal complication of DVT, has been reported to represent an increasing proportion of total venous thromboembolism events in elderly patients. Most pulmonary emboli originate from the femoropopliteal and pelvic veins. Multiple diagnostic tests are used for the diagnosis of thromboembolic disease (e.g., sonography, ventilation—perfusion scan, helical CT, or conventional pulmonary angiography). Recently, a comprehensive test, CT pulmonary angiography and venography, which can evaluate both DVT and pulmonary embolus, has been proposed to replace two or more separate examinations. Investigators have reported the feasibility and accuracy of combined CT pulmonary angiography and venography [4,5,6]. To our knowledge, difficulties and potential pitfalls in interpreting CT venography images that result in interobserver agreement variability have not been reported. This study was performed to evaluate interobserver variability in interpretation of CT venography images acquired after CT pulmonary angiography in patients suspected of having pulmonary embolism.

Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Patients
Between March 3, 1999 and April 28, 2000, 159 consecutive patients referred for helical CT to evaluate clinically suspected pulmonary embolism were considered for CT venography. Five patients for whom findings on leg sonography were negative for DVT within 24 hr before CT did not undergo CT venography. Eight patients with clinically suspected chronic pulmonary embolus and no leg symptoms also did not undergo CT venography. The remaining 146 patients who underwent CT venography in addition to CT pulmonary angiography formed the study group. Study patients included 131 men and 15 women (age range, 34-83 years; mean age, 61.74 years). Significant comorbid conditions were noted. All patients signed a written consent form approved by the hospital human subject committee and institutional review board.

CT
CT was performed on a model 5000 CT unit (Picker International, Cleveland, OH). The CT pulmonary angiography protocol that we used has been described [7]. Iodinated contrast medium was administered as a bolus with an automated injector (MCT Plus; Medrad, Pittsburgh, PA). A total of 100-150 mL of undiluted iopramide solution (Ultravist 300; Berlex Laboratories, Wayne, NJ) or iothalamate meglumine (Conray 60; Mallinckrodt Medical, St. Louis, MO) was injected at a rate of 4 mL/sec with a delay of 15-20 sec before scanning. In 25 patients who could not breath-hold, CT images were acquired during shallow breathing. Three to four minutes after the start of the injection, 5-mm-thick nonhelical axial CT venography images were acquired at 20-mm intervals (instead of the 50-mm intervals used in a recent study [4]) from the knees to the upper third of the abdomen. A 2-cm slice interval allowed the appropriate anatomic coverage in a reasonable time and with a limited number of images without skipping long venous segments. CT venography was performed with two different protocols: protocol 1, with which CT venography images were acquired with 100 mL of contrast medium and a 3-min delay, and protocol 2, with which a larger volume of contrast medium (150 mL) and longer delay (4 min) were used. Protocols 1 and 2 were used in 74 and 72 patients, respectively. Patients in both groups had similar demographic characteristics.

Leg Sonography
Bilateral leg sonography was performed with a 5- or 7-MHz linear array transducer (Logic 700, General Electric Medical Systems, Milwaukee, WI; or HDI 3000, Advanced Technology Laboratories, Bothell, WA) by experienced staff sonographers. Diagnostic criteria for acute DVT were lack of complete lumen obliteration with compression, flow void on color Doppler images, and lack of flow detection on spectral analysis [8]. The presence of venous expansion helped confirm the finding of acute DVT. Chronic DVT, as defined by vessel noncompressibility, wall thickening, and luminal contraction with or without flow on spectral analysis and color Doppler images, was not considered a positive study result for statistical analysis purposes. Valsalva's maneuver was routinely performed to assess valvular incompetence. Leg sonography performed within 24 hr before or after CT venography in 106 (74%) of 144 patients was the standard of reference for CT venography.

Image Analysis
All CT venography images were interpreted on three occasions. They first were interpreted prospectively, without knowledge of sonographic findings, at the time of examination after viewing images on the monitor or workstation and the hard copies. All CT venography images were then independently reviewed at a later date by a second interpreter (a body imaging fellow or an attending physician), who was also unaware of the sonographic findings. The observers first independently and subsequently in consensus completed a standard grading sheet. CT venography images (window width, 200-300 H; window level, 40-80 H) were graded for quality of venous opacification. Image quality was considered good when a high degree of venous opacification was shown, satisfactory when the images were sufficient for analysis of venous patency without a high degree of opacification, and poor when venous patency could not be assessed. Thirteen venous segments were analyzed in each patient (inferior vena cava, renal veins, iliac veins, common femoral veins, superficial femoral veins, proximal deep femoral or profunda veins, and popliteal veins).

The criterion used to diagnose acute DVT was the presence of a definite intraluminal filling defect as a primary finding and distention of the vein as a secondary finding (Fig. 1). The number of slices showing DVT was noted for CT venograms with positive findings for DVT. Findings of chronic DVT, such as calcified thrombi or venous walls, heterogenously enhancing small thick-walled veins, and presence of collateral veins, were again considered as negative findings for DVT, similar to sonography, for the purposes of statistical analysis.

View larger version (77K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1. —79-year-old man with acute deep venous thrombosis. Transverse CT venogram shows intraluminal filling defect in right common femoral vein (arrow), which is distended. Both observers interpreted this defect as evidence of deep venous thrombosis.

Criteria for the diagnosis of pulmonary emboli on CT pulmonary angiography have been previously described [9]. Two observers independently interpreted CT pulmonary angiography studies. In 25 CT pulmonary angiograms only central arteries were analyzable. Only two patients with clinical findings highly suggestive of pulmonary embolus who had inconclusive findings on CT pulmonary angiography underwent conventional pulmonary angiography.

Statistical Analysis
Interobserver variability was calculated as a kappa value. Strength of agreement was considered fair for kappa values of 0.21-0.40, moderate for values of 0.41-0.60, and good for values of 0.61-0.80 [10]. The level of a statistically significant difference was set at a p value of less than 0.05. All analyses were performed using the SAS statistical software package, version 6.12 (SAS Institute, Cary, NC).

Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Results are summarized in Tables 1 and 2. Many patients had known peripheral artery disease and most CT venographic studies revealed peripheral arterial calcifications. Thirty-five (24%) of 144 patients had known severe cardiopulmonary disease, including emphysema, pulmonary arterial hypertension, or cardiomyopathy. In 10 (29%) of these 35 patients, there was only satisfactory opacification of veins (Fig. 2). On a per-patient basis, 144 CT venography studies (99%) were graded as good or satisfactory. In two patients, CT venography studies were rated as poor and were not considered for further analysis. A total of 1898 venous segments were analyzed in 144 patients with 146 CT venograms. Of 312 (16%) of 1898 venous segments that were not analyzable, 206 venous segments (66%) were not analyzable because of poor opacification and 42 because of beam hardening artifacts from orthopedic hardware or dense arterial calcifications. Other reasons for the nonanalyzable venous segments included indwelling vascular catheters (n = 10), small nearly collapsed veins in patients with below-the-knee amputations (n = 13), and possible physiologic compression of the superficial femoral vein in the mid to distal thigh (n = 41). Only short segments of calf veins were examined in our study. Evaluation of the calf veins was not used for statistical analysis. There was no DVT seen in the inferior vena cava or renal veins in any patient. In three patients, DVT was seen on only one slice; in other patients, DVT was seen on two or more slices.

View larger version (73K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2. —66-year-old man with suboptimal opacification of veins. Transverse CT venogram shows apparent intraluminal filling defect in right iliac vein (arrow), which was interpreted as positive for deep venous thrombosis by one observer and negative by second observer. Consensus interpretation and sonography were negative for deep venous thrombosis.

Of 146 CT venograms, disagreement occurred in 17 (12%). In 13 examinations, the disagreement occurred at the level of the profunda, mid superficial femoral vein, or the pelvic veins (Fig. 2); in two, the disagreement occurred in the interpretation of the first slice at the knee (Fig. 3A,3B); and in two, chronic DVT was interpreted as acute (Fig. 4A,4B). Of the 17 CT venograms with interobserver agreement variability, 11 were finally interpreted as negative and six as positive for DVT at consensus interpretation. Fifteen of these 17 patients had sonographic images available for comparison. Consensus reading was false-negative in one patient and false-positive in three patients when findings of CT venography were compared with sonographic findings. Of the 17 CT venograms, 14 showed only satisfactory venous opacification and three showed good, homogeneous opacification.

View larger version (104K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3A. —69-year-oldman with deep venous thrombosis seen only on one slice. First transverse CT venogram slice at level of knee joint line shows intraluminal filling defect (arrow) in left lesser saphenous vein at confluence with popliteal vein. Defect was missed by one observer.

View larger version (125K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3B. —69-year-old man with deep venous thrombosis seen only on one slice. Transverse sonogram at same level as A shows intraluminal echogenic material distending lesser saphenous vein (arrows). Involved vein was not compressible and deep venous thrombosis extended 4-5 cm below knee.

View larger version (55K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4A. —75-year-old man with chronic deep venous thrombosis. Transverse CT venogram at inguinal ligament shows nonoccluding filling defect along posterior aspect of right common femoral vein (arrow). Defect was interpreted as evidence of acute deep venous thrombosis by both observers.

View larger version (141K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4B. —75-year-old man with chronic deep venous thrombosis. Longitudinal color Doppler sonogram of right common femoral vein shows thickening of posterior wall of right common femoral vein (arrowheads) consistent with chronic deep venous thrombosis. This finding was unchanged from previous sonogram obtained 6 months earlier (not shown).

CT pulmonary angiographic findings were positive for pulmonary embolus in 23 patients (16%). Overall agreement for CT pulmonary angiography was 96% ( = 0.85) and agreement for protocols 1 and 2 was 95% ( = 0.81) and 97% ( = 0.89), respectively.

Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
We found moderately good interobserver agreement for the diagnosis of acute DVT using CT venography performed in addition to CT pulmonary angiography in patients suspected of having pulmonary embolism. To our knowledge, there are no previous studies that have assessed the interobserver variability for CT venography; however, agreement rates for CT pulmonary angiography have been reported.

Interobserver agreement in our cohort of patients for CT pulmonary angiography was 96%, which is similar to the high range of values (75-96%) reported previously [11,12,13]. This was better than the agreement for CT venography. Although the pulmonary artery system anatomy is more complex compared with that of the deep venous system, we believe that one of the main reasons for the difference in agreement rates between CT venography and CT pulmonary angiography is a result of the relatively poor opacification of the veins and more flow-related artifacts on CT venography in our patients.

Many of our study patients had peripheral vascular disease that probably resulted in poor delivery of contrast bolus to the deep venous system, which may have caused the relatively poor opacification of the veins. Therefore, during the course of our study, we hypothesized that a larger volume of contrast medium and a longer delay may result in better opacification. However, this did not result in a statistically significant improvement in opacification of the veins or in agreement rate, which may indicate more complex hemodynamic factors due to a combination of abnormal cardiac contractility and peripheral arterial disease. It is difficult to be certain whether more familiarity with the interpretation of CT pulmonary angiography findings compared with that of CT venography findings or whether the results of CT pulmonary angiography had any impact on the interobserver agreement rate for CT venography in our study.

Other pitfalls in interpretation of CT venograms encountered in this study were caused by beam-hardening artifacts from densely calcified arteries or arterial thrombi, or artifacts from orthopedic hardware. In one patient, arterial thrombus was misinterpreted as DVT; however, arterial thrombus can easily be differentiated from venous thrombus if the veins are systematically analyzed from one image to the next.

Interobserver variability was more common in the interpretation of the pelvic veins, the profunda femoris vein, and the mid superficial femoral venous segments. The course of pelvic veins, which in some segments is not perpendicular to the imaging plane, probably results in volume averaging, which could potentially result in a pseudofilling defect. Asymmetry of the caliber of the profunda vein with flow-related artifacts in the varicose vein resulted in an apparent filling defect in two patients. The inherent compression of the superficial femoral vein caused by adjacent muscles and the position of the legs on the CT table may partially explain the relatively poor visualization of the mid superficial femoral vein in some patients compared with the common femoral vein and popliteal veins, which are generally surrounded by fat.

In two patients the interobserver disagreement occurred when only one image (the first image at the knee) revealed a filling defect (Fig. 3A,3B). Sonography in both patients showed venous thrombus extending inferiorly in the deep calf veins (Fig. 3A,3B). In another patient, DVT in the common femoral vein was seen on only one slice. Although a recent study [4] reported 100% correlation between CT venography and sonographic findings despite use of a 5-cm slice interval, the results of our study indicate that using a slice interval greater than 2 cm can potentially lead to either false-negative findings on CT venography or an underestimation of the extent of thromboembolic disease. It can be argued that if the slice interval had been narrower and if the imaging of legs had started at least 2 cm below the knee joint line resulting in multiple slices revealing DVT, the agreement could have been better. Future studies are warranted with different techniques to assess if narrower intervals, especially at the knee, groin, and pelvis, where DVT commonly occurs, and imaging of the upper calves will improve the agreement rate for interpretation of CT venography studies.

Observers became familiar with the findings of chronic DVT on CT venography only during the course of this study, and differentiation of acute from chronic DVT was not our main objective. We learned that calcified thrombi and veins and collateral veins may be better seen on CT venography; however, wall thickening, webs, and valvular incompetence are usually better assessed with sonography. Rarely, a nonoccluding filling defect caused by focal wall thickening can mimic an acute DVT on CT venography (Fig. 4A,4B). In clinical practice the differentiation between acute and chronic DVT is important for determination of the need and duration of anticoagulant therapy. It is also important when the safety and efficacy of a new thrombolytic or anticoagulant therapy is being investigated [14].

There are limitations to our study. There was no correlation with sonography in one third of our patients and no correlation with the suprainguinal veins; however, the accuracy of the technique was not primarily assessed in this study. Although sonography or other confirmatory tests were not available in all 17 CT venography studies in which interobserver disagreement occurred, we believe that consensus reading in difficult cases may help to avoid a false-positive interpretation.

In conclusion, interobserver agreement for CT venography studies that are negative for DVT is moderately good. Optimization of technical factors, such as the thinner slices at the narrower intervals that are possible with multidetector CT, to improve anatomic coverage and spatial resolution may result in better agreement. Good homogeneous opacification is the most important factor in consistent reporting. Although a diagnostic-quality opacification can be achieved in most patients, in a few patients the opacification is not optimal even with a longer delay or larger volume of contrast medium.

References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O'Fallon WM, Melton LJ. Trends in the incidence of deep venous thrombosis and pulmonary embolism: a 25-year population-based study. Arch Intern Med 1998;158:585 -593[Abstract/Free Full Text]
2. Gillum RF. Pulmonary embolism and thrombophlebitis in the United States, 1970-1985. Am Heart J 1987;114:1262 -1264[Medline]
3. Anderson FA Jr, Wheeler HB, Goldberg RJ, et al. A population-based perspective of the hospital incidence and case-fatality rates of deep venous thrombosis and pulmonary embolism: the Worcester DVT Study. Arch Intern Med 1991;151:933 -938[Abstract/Free Full Text]
4. Loud PA, Katz DS, Klippenstein DL, Shah RD, Grossman ZD. Combined CT venography and pulmonary angiography in suspected thromboembolic disease: diagnostic accuracy for deep venous evaluation. AJR 2000;174:61 -65[Abstract/Free Full Text]
5. Yankelevitz DF, Gamsu G, Shah A, et al. Optimization of combined CT pulmonary angiography with lower extremity CT Venography. AJR 2000;174:67 -69[Abstract/Free Full Text]
6. Garg K, Kemp JL, Wojcik D, et al. Thromboembolic disease: prospective comparison of combined CT pulmonary angiography and venography with bilateral leg sonography in 70 patients. AJR 2000;175:997 -1001[Abstract/Free Full Text]
7. Garg K, Welsh CH, Feyerabend AJ, et al. Pulmonary embolism: diagnosis with spiral CT and ventilation-perfusion scanning—correlation with pulmonary angiographic results or clinical outcome. Radiology 1998;208:201 -208[Abstract/Free Full Text]
8. Cronan JJ. Venous thromboembolic disease: the role of US. Radiology 1993;186:619 -630[Abstract/Free Full Text]
9. Remy-Jardin M, Remy J, Deschildre F, et al. Diagnosis of pulmonary embolism with spiral CT: comparison with pulmonary angiography and scintigraphy. Radiology 1996;200:699 -706[Abstract/Free Full Text]
10. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159 -174[Medline]
11. Goodman LG, Curtin JJ, Mewissen MW, et al. Detection of pulmonary embolism in patients with unresolved clinical and scintigraphic diagnosis: helical CT versus angiography. AJR 1995;164:1369 -1374[Abstract/Free Full Text]
12. Mayo JR, Remy-Jardin M, Muller NL, et al. Pulmonary embolism: prospective comparison of spiral CT with ventilation-perfusion scintigraphy. Radiology 1997;205:447 -452[Abstract/Free Full Text]
13. Lefebvre CC, Howarth N, Lucidarme O, et al. Contrast-enhanced helical CT for pulmonary embolism detection: inter- and intraobserver agreement among radiologists with variable experience. AJR 1999;172:107 -112[Abstract/Free Full Text]
14. The Columbus Investigators. Low-molecular-weight heparin in the treatment of patients with venous thromboembolism. N Engl J Med 1997;337:657 -662[Abstract/Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				H. Nazaroglu, C. A. Ozmen, H. O. Akay, I. Kilinc, and A. Bilici 64-MDCT Pulmonary Angiography and CT Venography in the Diagnosis of Thromboembolic Disease Am. J. Roentgenol., March 1, 2009; 192(3): 654 - 661. [Abstract] [Full Text] [PDF]

				L. R. Goodman, H. D. Sostman, P. D. Stein, and P. K. Woodard CT Venography: A Necessary Adjunct to CT Pulmonary Angiography or a Waste of Time, Money, and Radiation? Radiology, February 1, 2009; 250(2): 327 - 330. [Full Text] [PDF]

				L. R. Goodman, P. D. Stein, F. Matta, H. D. Sostman, T. W. Wakefield, P. K. Woodard, R. Hull, D. F. Yankelevitz, and A. Beemath CT Venography and Compression Sonography Are Diagnostically Equivalent: Data from PIOPED II Am. J. Roentgenol., November 1, 2007; 189(5): 1071 - 1076. [Abstract] [Full Text] [PDF]

				M. Remy-Jardin, M. Pistolesi, L. R. Goodman, W. B. Gefter, A. Gottschalk, J. R. Mayo, and H. D. Sostman Management of Suspected Acute Pulmonary Embolism in the Era of CT Angiography: A Statement from the Fleischner Society Radiology, November 1, 2007; 245(2): 315 - 329. [Full Text] [PDF]

				L. R. Goodman, P. D. Stein, A. Beemath, H. D. Sostman, T. W. Wakefield, P. K. Woodard, and D. F. Yankelevitz CT Venography for Deep Venous Thrombosis: Continuous Images Versus Reformatted Discontinuous Images Using PIOPED II Data Am. J. Roentgenol., August 1, 2007; 189(2): 409 - 412. [Abstract] [Full Text] [PDF]

				H. Arakawa, T. Kohno, T. Hiki, and Y. Kaji CT Pulmonary Angiography and CT Venography: Factors Associated with Vessel Enhancement Am. J. Roentgenol., July 1, 2007; 189(1): 156 - 161. [Abstract] [Full Text] [PDF]

				B. Ghaye, A. Nchimi, C. T. Noukoua, and R. F. Dondelinger Does Multi-Detector Row CT Pulmonary Angiography Reduce the Incremental Value of Indirect CT Venography Compared with Single-Detector Row CT Pulmonary Angiography? Radiology, July 1, 2006; 240(1): 256 - 262. [Abstract] [Full Text] [PDF]

				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]

				M. D. Cham, D. F. Yankelevitz, and C. I. Henschke Thromboembolic Disease Detection at Indirect CT Venography versus CT Pulmonary Angiography Radiology, February 1, 2005; 234(2): 591 - 594. [Abstract] [Full Text] [PDF]

				D. S. Katz, P. A. Loud, D. Bruce, A. M. Gittleman, R. Mueller, D. L. Klippenstein, and Z. D. Grossman Combined CT Venography and Pulmonary Angiography: A Comprehensive Review RadioGraphics, October 1, 2002; 22(90001): S3 - 19. [Abstract] [Full Text] [PDF]

				J. R. Mayo, L. H. Ketai, P. A. Loud, Z. D. Grossman, D. S. Katz, and A. M. Gittleman Invited Commentary * Authors' Response RadioGraphics, October 1, 2002; 22(90001): S20 - 24. [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited 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 Garg, K. Articles by Barón, A. E. Search for Related Content PubMed PubMed Citation Articles by Garg, K. Articles by Barón, A. E. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?