| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Thoracic Imaging |
1 From the Departments of Clinical Radiology (U.J.S., T.H., N.H., R.D.B., C.R.B., M.F.R.), Biometry and Epidemiology (S.A.), and Internal Medicine (O.M., A.K., R.H.), Klinikum Grosshadern, University of Munich, Marchioninistr 15, 81377 Munich, Germany, and the Department of Diagnostic Radiology, Kyungpook National University Hospital, Taegu, Korea (D.S.K.). From the 1998 RSNA scientific assembly. Received August 6, 1998; revision requested September 4; final revision received May 10, 1999; accepted August 20. Address reprint requests to U.J.S. (e-mail: schoepf@ikra.med.uni-muenchen.de).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: CT angiography of the pulmonary arteries was performed in 56 patients to rule out pulmonary embolism. Electron-beam CT was performed in 28 patients. The other 28 patients underwent spiral CT with comparable scanning protocols. The depiction of segmental and subsegmental arteries was analyzed by three independent readers. The contrast enhancement in the main pulmonary artery was measured in each patient.
RESULTS: Analysis was performed in 1,120 segmental and 2,240 subsegmental arteries. One segmental (RA7, P = .010) and two subsegmental (LA7b, P = .029; RA6a+b, P = .038) arteries in paracardiac and basal segments of the lung were depicted significantly better with electron-beam CT. There was no statistically significant difference between electron-beam and spiral CT in the total number of analyzable peripheral arteries depicted. The mean contrast enhancement in the main pulmonary artery was 362 HU in electron-beam CT studies versus 248 HU in spiral CT studies.
CONCLUSION: Detailed visualization of peripheral pulmonary arteries is well within the scope of advanced CT techniques. Electron-beam CT has minor advantages in analyzing paracardiac arteries, probably because of reduction of motion artifacts and higher contrast enhancement. Further studies are needed to establish whether electron-beam CT allows a more confident diagnosis of emboli in these vessels.
Index terms: Computed tomography (CT), comparative studies, 944.12915, 944.12919 • Computed tomography (CT), contrast enhancement, 944.12915, 944.12919 • Computed tomography (CT), electron beam, 944.12919 • Computed tomography (CT), helical, 944.12915 • Embolism, pulmonary, 60.721, 944.77 • Pulmonary arteries, CT, 944.12915, 944.12919
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The reliability of CT in the detection of smaller clots in segmental or subsegmental arteries, however, is still debated (7–10). Electron-beam CT has theoretic advantages over spiral CT in the detection of such peripheral emboli. It is argued that electron-beam CT leads to a reduction of motion artifact by means of its shorter acquisition times and allows scanning during maximal opacification of vessels (11,12). However, to our knowledge, the value of electron-beam CT has never been compared with that of spiral CT in the evaluation of pulmonary arteries.
In a carefully designed study, Remy-Jardin et al (13) recently chose a patient population in whom PE had been excluded so that they could test for the effect of collimation on the evaluation of the pulmonary vessels by means of spiral CT, without double exposure to the patients. We adapted a similar strategy for the purpose of this study, which was to compare the value of electron-beam CT with that of spiral CT in the analysis of segmental and subsegmental pulmonary arteries; on the basis of this approach, we present a comparison of the two best-documented (4,6–8) CT techniques for the diagnosis of PE.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Patients were included in the study if the following criteria were met: (a) PE had been ruled out confidently because the chest CT scan was negative for PE (all patients). The CT diagnosis had been confirmed by means of either a normal pulmonary angiogram (n = 8; group A: n = 3, group B: n = 5) or a normal ventilation-perfusion scan negative for PE (n = 48; group A: n = 25, group B: n = 23) within 24 hours after the CT investigation. A normal ventilation-perfusion scan is generally accepted to rule out PE (14–16); therefore, ventilation-perfusion scanning is used frequently in our institution to exclude peripheral PE definitely after central embolism and alternative disease are ruled out at CT. For interpretation, Prospective Investigation of Pulmonary Embolism Diagnosis, or PIOPED, classifications were used (15). (b) There was no pulmonary condition that could compromise the evaluation of the peripheral vessels (ie, a history of lung surgery, pulmonary metastases, or infiltrates). (c) There was no suspicion of pulmonary hypertension (ie, known history of pulmonary hypertension or dilatation of the right side of the heart or central pulmonary arteries). (d) Data acquisition had been achieved in a single breath hold.
The chest CT scans that were obtained in the first 28 consecutive patients with each scanner and that fulfilled these criteria were included in the study. By including only patients whose pulmonary arteries could be expected to be normal, we intended to create two normal populations that allowed the comparison of two scanning techniques without double exposure to the patients.
Scanning Protocols
For this study, our CT protocols for the diagnosis of PE were standardized as follows: All patients underwent caudocephalic scanning in a supine position and at end-inspiratory suspension during a single breath hold. Prior to each study, a localizing scan was obtained. The z axis coverage and the field of view were chosen to include the entire thorax to allow for complete evaluation of the lung parenchyma. With the electron-beam scanner, an overlap mode was used with 6-mm collimation and 6-mm section thickness, which resulted in a pitch of 1.18. Exposure time was 0.2 seconds at 130 kV and 640 mA. Overlapping transverse images were then reconstructed with a 3-mm interval. Depending on the dimensions of the patient, scanning times were 14–17 seconds (mean, 16 seconds).
On the spiral scanner, care was taken to keep the scanning parameters as comparable to those used with the electron-beam CT scanner as achievable, given the different technical concept of the two scanners. Along with 0.75 second per 360° rotation, we chose a collimation of 5 mm with a table feed of 9 mm for a pitch of 1.8, 120 kV, and 170 mA. With use of a 180o-linear interpolation algorithm, these parameters result in an effective section thickness of 6 mm (17). Overlapping transverse images were then reconstructed with a 3-mm interval by using a kernel comparable to the "very sharp" kernel of the electron-beam CT scanner (abdominal 70). Again, depending on the dimensions of the patient, the duration of data acquisition was 23–32 seconds (mean, 26 seconds).
All patients received intravenous injection of a bolus of 120 mL of iopentol (Imagopaque 300; Nycomed, Oslo, Norway) at a rate of 4 mL/sec. The contrast agent was administered via an 18-gauge venous access in a cubital or antecubital vein by using a power injector (group A: XD 5500, Ulrich, Ulm, Germany; group B: MCT-Plus-FL, Medrad, Pittsburgh, Pa). Since cardiac disease other than suspected pulmonary hypertension was not an exclusion criterion, start delays were individually adapted to each patient on the basis of criteria such as age, general condition, and heart size as seen on the localizing scan. Scanning delays were 12–19 seconds (mean, 16 seconds) in group A and 11–17 seconds (mean, 14 seconds) in group B. Automatic bolus triggering, which is available on the spiral scanner, was not used so that the injection protocols would be comparable and because we do not routinely use this technique for the purpose of PE diagnosis. Using the described protocols, we were able to achieve satisfactory opacification of the pulmonary arteries in all patients: Contrast enhancement was considered sufficient subjectively for the detection of filling defects (8,13).
Image Analysis
All images were obtained at identical mediastinal (width, 340 HU; level, 40 HU) and lung (width, 1,400 HU; level, -400 HU) window settings and were viewed on hard copies. Images were rated independently by three experienced radiologists (T.H., N.H., D.S.K.). We did not attempt to blind the readers to the scanning technique because of the obvious differences in image characteristics generated by the two scanners. To identify segmental and subsegmental arteries, we used the nomenclature as outlined by Remy-Jardin et al (13). This nomenclature is based on the standard descriptions by Jackson and Huber (18) and Boyden (19), with slight modifications to account for anatomic variations and the orientation of vessels in the transverse plane on CT scans. Twenty segmental and 40 subsegmental arteries are described in this nomenclature.
Prior to image analysis, a training session was held during which the readers were familiarized with the modifications in nomenclature and agreed on the following strategy for the analysis. An artery was considered analyzable when it was found in the expected, prevailing anatomic location and showed definite contrast enhancement without filling defects from its proximal to distal portions. Partial volume effects, defined as relative hypoattenuation due to the small size of an artery or its obliquity to the scanning plane, were not considered grounds for deeming an artery nonanalyzable in cases in which the partial volume effect was clearly caused by the branching of an artery or in cases in which the contrast agent–filled lumen continued as expected on images obtained at subsequent levels. We did not attempt to account for anatomic variants, which, if present, were considered nonanalyzable for the purpose of this study.
To test whether the rapid data acquisition of the electron-beam scanner allowed imaging during optimal opacification of contrast agent (11), one author (U.J.S.) drew equal-sized regions of interest at identical levels of the main pulmonary artery on all 56 scans on a digital workstation (Magic View 1000/VA02D; Siemens Medical Systems). The mean opacification within these regions of interest was recorded in Hounsfield units after background noise had been subtracted.
Statistical Analysis
The distribution-free Mann-Whitney U test was used to compare the total number of arteries that were coded as analyzable with the two scanning techniques. To compare the values for each segmental and subsegmental artery separately, Fisher exact tests for fourfold tables were performed on the pooled data from the three readers by calculating the mean percentage of arteries the three readers found analyzable in each vascular territory. To test for statistically significant differences at a local level, a P value less than .05 was considered to indicate a significant difference in each vascular territory. Interreader correlation was determined by comparing the total number of arteries that each reader found analyzable in the 20 segmental and 40 subsegmental vascular zones on electron-beam and spiral CT scans. An r value of more than 0.75 was considered to indicate good correlation between the readers. The Student t test was used to compare contrast enhancement in the main pulmonary artery and the descending aorta in electron-beam and spiral CT studies.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Number of Analyzable Segmental Arteries
The three readers analyzed 1,120 segmental arteries. When each segmental artery was considered separately, a statistically significant difference (P < .05) in the mean percentage of analyzable arteries between the two methods was found for the paracardiac segmental artery of the right lower lobe (RA7; electron-beam CT, 100%; spiral CT, 79%; P = .010) (Table 1).
|
For most segments, the mean percentage of analyzable segmental arteries was 80%–100% with either method. In the right lung, a frequency of less than 80% was observed for the posterior segmental artery of the right upper lobe (RA3) and the paracardiac segment of the right lower lobe (RA7) when the spiral technique was used. In the left lung, less than 80% of arteries were considered analyzable in the anterior segment (LA8; electron-beam CT, 82%; spiral CT, 75%) at spiral CT. With electron-beam CT, less than 80% of arteries were considered analyzable in the apical segment of the left upper lobe (LA1; electron-beam CT, 71%; spiral CT, 89%) and the laterobasal segment of the left lower lobe (LA9; electron-beam CT, 79%; spiral CT, 86%). Frequencies of less than 80% with both methods were found for the medial segment of the left lower lobe (LA7; electron-beam CT, 79%; spiral CT, 64%). On the segmental level, good interobserver agreement was achieved for spiral CT (r = 0.79) and electron-beam CT (r = 0.83).
Number of Analyzable Subsegmental Arteries
The three readers analyzed 2,240 subsegmental arteries. When each subsegment was considered separately, statistically significant differences in the mean percentage of analyzable arteries between the two methods were found for the medial subsegmental artery of the paracardiac segment of the left lower lobe (LA7b; electron-beam CT, 75%; spiral CT, 46%; P = .029) and for the superomedial subsegmental branch in the apical segment of the right lower lobe (RA6a+b; electron-beam CT, 100%; spiral CT, 86%; P = .038). Slight differences above a significance level of .05 were recognizable for the posterior subsegment of the upper lingular division (LA4a; electron-beam CT, 68%; spiral CT, 57%) (Table 2).
|
Twenty-three of 40 subsegmental arteries were considered adequately depicted with a frequency of more than 70% with both methods (Table 2). The vascular zones that were considered analyzable in less than 70% of cases with one of the methods show a clear pattern: The lateral rami of the anterior and posterior segmental arteries in both upper lobes (RA2a, RA3a, LA2a, and LA3a), all subsegmental arteries in the middle lobe and lingula (RA4a, RA4b, RA5a, RA5b, LA4a, LA4b, LA5a, LA5b), and the subsegmental rami in the paracardiac segments in both lower lobes (RA7a, RA7b, LA7a, LA7b) were considered not adequately depicted in more than 30% of cases with at least one of the two techniques. LA9a was considered analyzable in 68% of cases with electron-beam CT (Table 2).
A frequency of analyzable arteries of more than 70% was found by using electron-beam CT and of less than 70% by using spiral CT for RA3a, RA5a, RA7a, RA7b, LA5a, and LA7b. A frequency of analyzable arteries of more than 70% at spiral CT and less than 70% at electronbeam CT was found for LA2a. On the subsegmental level, good or moderate interobserver agreement was achieved (electron-beam CT, r = 0.76; spiral CT, r = 0.71).
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Comparing different scanning techniques in the same patient with documented PE is hampered by ethical concerns, such as double exposure and delay of treatment, and may also be compromised by pitfalls in the interpretation of studies, such as those stemming from changes in clot formations and autofibrinolysis in a dynamic disease process. By comparing different scanning techniques in two normal populations, we avoided such problems. Using this approach, we were able to show that an accurate visualization of the peripheral arterial bed of the lung is within the scope of advanced CT scanning techniques.
Vascular zones that could not be optimally visualized included the posterior right and the apical left upper lobe. Inadequate depiction of these segmental arteries was reported in previous studies (7,13) and was attributed to the short axis of RA3 and the anatomic variability of LA1 (13). Poor visualization was also reported for the paracardiac segments of the left lower lobe (13). These findings correspond well to the less adequate depiction of LA7, LA8, and RA7 in our study. The significantly better depiction of RA7 in the electron-beam CT studies makes it especially likely that transmitted cardiac pulsations contribute to the relatively poor visualization of these vessels, which may be amenable to the short scanning times of electron-beam CT.
On the subsegmental level, the less adequately depicted vessels in the upper lobes were the lateral rami of RA2, RA3, LA2, and LA3. Poor visualization of upper lobe arteries was also found in the preceding studies and was explained by inadequate contrast enhancement (7) and anatomic variance, particularly of the lateral rami (13). Small size (13), anatomic variability (13), and obliquity to the transverse scanning plane (7) were explanatory of the suboptimal depiction of the middle lobe and lingular vessels, which were also observed in our study.
Yet another factor precluding the adequate visualization of the lingular vessels in previous studies may be their proximity to the moving heart, if one considers the slightly better depiction of LA4a in the patients who underwent electron-beam CT in our study. Blurring of the image owing to cardiac motion artifacts has already been recognized as impairing the proper analysis of the left paracardiac subsegmental arteries (13). In our study, these rami were also significantly (LA7b, P = .029) better depicted at electron-beam CT, likely because of the suppression of cardiac motion artifacts by the shorter scanning times of electron-beam CT. The difference in the analysis of the apical subsegmental artery of the right upper lobe, finally, may be attributable to better contrast agent opacification of vessels at electron-beam CT.
It has been argued that electron-beam CT short scanning times allow image acquisition during peak contrast enhancement (11,12). The significantly higher contrast enhancement achieved in electron-beam CT examinations in our study corresponds well to this notion. Besides the suppression of cardiac motion artifacts at electron-beam CT, better contrast agent opacification of vessels thus may also serve to explain the differences between electron-beam CT and spiral CT in the depiction of some vessels.
There are certain limitations to this study that need to be mentioned. Since we compared two normal populations, we could not account for anatomic variants and assumed that their occurrence would be evenly distributed in our study population. Limitations also arise from the different technical concepts of the two scanners. Therefore, care was taken to keep the protocols comparable without compromising the strength of each scanner. The published protocols for the diagnosis of PE with electron-beam CT use 6-mm collimation (6,7). Thinner collimations (ie, 2 mm) result in an improvement in the analysis of segmental and subsegmental pulmonary arteries (13). However, the current thermal limitation of the electron-beam scanner—a maximal scanning time of 17 seconds—does not allow sufficient z axis coverage with thinner collimations and increments in tall patients. Therefore, the scanning parameters on both scanners were set to result in an effective section thickness of 6 mm (17), which allows 3-mm reconstructions from overlapping images.
For the evaluation, we preferred individual reading, since independent evaluation by several experienced radiologists was considered to be better suited to detect real differences between the two techniques more powerfully than would consensus reading.
The clinical need for a meticulous analysis of the peripheral pulmonary vessels is currently being discussed. It has been shown that 6%–30% of patients with documented PE have clots in only subsegmental and smaller arteries (15,20). There has been a degree of uncertainty, however, about the clinical importance of such small peripheral clots in the absence of central emboli. It is assumed that one important function of the lung is to prevent small emboli from entering the arterial circulation (16). Such emboli are thought to form even in healthy persons, although to our knowledge this notion has never been substantiated (21). Controversy also exists as to whether the treatment of small emboli, once detected, may result in a better clinical outcome for patients (8,16,20,22).
There seems to be agreement, though, that the presence of peripheral emboli is an important indicator for current deep venous thrombosis, which thus potentially heralds more severe embolic events (1,20,23). A burden of small peripheral emboli may also have prognostic relevance for persons with cardiopulmonary restrictions (16,20) and for the development of chronic pulmonary hypertension in patients with thromboembolic disease (20). Before the argument over the clinical importance of small clots is finally settled, an accurate analysis of the segmental and subsegmental pulmonary arteries appears to be justified.
For the segmental level and for the analysis of subsegmental vessels with electron-beam CT, good interobserver correlation was achieved (r = 0.76–0.83). For the analysis of subsegmental vessels with spiral CT, interobserver correlation was somewhat lower (r = 0.71), possibly because of lower opacification and motion artifacts that led to a more divergent assessment of subsegmental arteries by the readers. However, this slightly lower interobserver correlation at spiral CT needs to be seen in the light of recent data that show that at pulmonary angiography initial interobserver agreement in the individual interpretation of subsegmental vessels can be achieved in only 45% of cases (24).
Despite electron-beam CT shorter scanning times, which result in sharper delineation and better contrast enhancement of pulmonary arteries (Figs 1–4), our study revealed only minor advantages of electron-beam CT over spiral CT in some segments and subsegments of the pulmonary arterial system. This may be attributable to image noise due to electron-beam CT short scanning times and the fact that spiral CT is already a rather efficient tool for the visualization of segmental and subsegmental pulmonary arteries. In addition, electron-beam CT is available at only a limited number of centers specializing in cardiac care. The excellent results for the analysis of segmental and subsegmental arteries that were achieved with both modalities, however, emphasize the benefit to be gained if CT is included in the diagnostic algorithm for PE.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, U.J.S., M.F.R.; study concepts, U.J.S., C.R.B., M.F.R.; study design, U.J.S., D.S.K.; definition of intellectual content, U.J.S., R.D.B., M.F.R.; literature research, U.J.S.; clinical studies, U.J.S., C.R.B., O.M., A.K., R.H., D.S.K.; data acquisition, U.J.S., C.R.B., O.M., A.K., R.H.; data analysis, T.H., N.H., R.D.B., D.S.K.; statistical analysis, S.A., U.J.S.; manuscript preparation and editing, U.J.S.; manuscript review, C.R.B., N.H., T.H., M.F.R., D.S.K.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. K. Yousefzadeh, M. B. Ward, and C. Reft Internal Barium Shielding to Minimize Fetal Irradiation in Spiral Chest CT: A Phantom Simulation Experiment. Radiology, June 1, 2006; 239(3): 751 - 758. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. G. Flohr, S. Schaller, K. Stierstorfer, H. Bruder, B. M. Ohnesorge, and U. J. Schoepf Multi-Detector Row CT Systems and Image-Reconstruction Techniques Radiology, June 1, 2005; 235(3): 756 - 773. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Reinartz, J. E. Wildberger, W. Schaefer, B. Nowak, A. H. Mahnken, and U. Buell Tomographic Imaging in the Diagnosis of Pulmonary Embolism: A Comparison Between V/Q Lung Scintigraphy in SPECT Technique and Multislice Spiral CT J. Nucl. Med., September 1, 2004; 45(9): 1501 - 1508. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. J. Schoepf, S. Z. Goldhaber, and P. Costello Spiral Computed Tomography for Acute Pulmonary Embolism Circulation, May 11, 2004; 109(18): 2160 - 2167. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. J. Schoepf and P. Costello CT Angiography for Diagnosis of Pulmonary Embolism: State of the Art Radiology, February 1, 2004; 230(2): 329 - 337. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Coche, F. Verschuren, A. Keyeux, P. Goffette, L. Goncette, P. Hainaut, F. Hammer, E. Lavenne, F. Zech, P. Meert, et al. Diagnosis of Acute Pulmonary Embolism in Outpatients: Comparison of Thin-Collimation Multi-Detector Row Spiral CT and Planar Ventilation-Perfusion Scintigraphy Radiology, December 1, 2003; 229(3): 757 - 765. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Tillie-Leblond, I. Mastora, F. Radenne, S. Paillard, A.-B. Tonnel, J. Remy, and M. Remy-Jardin Risk of Pulmonary Embolism after a Negative Spiral CT Angiogram in Patients with Pulmonary Disease: 1-year Clinical Follow-up Study Radiology, May 1, 2002; 223(2): 461 - 467. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U J Schoepf, C R Becker, R D Bruening, B M Ohnesorge, A Huber, L-G Haw, H Hildebrandt, and M F Reiser Multislice CT angiography Imaging, December 15, 2001; 13(5): 357 - 365. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Ghaye, D. Szapiro, I. Mastora, V. Delannoy, A. Duhamel, J. Remy, and M. Remy-Jardin Peripheral Pulmonary Arteries: How Far in the Lung Does Multi-Detector Row Spiral CT Allow Analysis? Radiology, June 1, 2001; 219(3): 629 - 636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. C. Velmahos, P. Vassiliu, A. Wilcox, S. E. Hanks, A. Salim, D. Harrel, S. Palmer, and D. Demetriades Spiral Computed Tomography for the Diagnosis of Pulmonary Embolism in Critically Ill Surgical Patients: A Comparison With Pulmonary Angiography Arch Surg, May 1, 2001; 136(5): 505 - 510. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. J. Schoepf, R. Bruening, H. Konschitzky, C. R. Becker, A. Knez, J. Weber, O. Muehling, P. Herzog, A. Huber, R. Haberl, et al. Pulmonary Embolism: Comprehensive Diagnosis by Using Electron-Beam CT for Detection of Emboli and Assessment of Pulmonary Blood Flow Radiology, December 1, 2000; 217(3): 693 - 700. [Abstract] [Full Text]
|
|
|
|
 |
|
 C. H. McCollough, M. R. Bruesewitz, T. R. Daly, and F. E. Zink Motion Artifacts in Subsecond Conventional CT and Electron-Beam CT: Pictorial Demonstration of Temporal Resolution RadioGraphics, November 1, 2000; 20(6): 1675 - 1681. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. J. Schoepf, N. Holzknecht, T. K. Helmberger, A. Crispin, C. Hong, C. R. Becker, and M. F. Reiser Subsegmental Pulmonary Emboli: Improved Detection with Thin-Collimation Multi-Detector Row Spiral CT Radiology, February 1, 2002; 222(2): 483 - 490. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||
