| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Thoracic Imaging |
1 From the Department of Radiology, Kobe University Graduate School of Medicine, 7-5-2 Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan (Y.O., D.T., S.A., K.S.); Department of Radiology, Pulmonary Functional Imaging Research, University of Pennsylvania Medical Center, Philadelphia (Y.O., H.H.); and Hyogo Medical Center for Adults, Akashi, Japan (M.K.). Received June 4, 2001; revision requested June 28; revision received November 13; accepted January 8, 2002. Address correspondence to Y.O. (e-mail: yosirad@kobe-u.ac.jp).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Fifty-eight patients with 58 pathologic analysis–proved SPNs (diameter < 30 mm) underwent dynamic 1.5-T MR imaging. The 58 SPNs were classified into three groups at pathologic analysis: malignant SPNs (n = 38), active infections (n = 10), or benign SPNs (n = 10). From signal intensity–time curves generated after the bolus injection of contrast material, the maximum relative enhancement ratio and slope of enhancement were calculated and statistically compared among the three groups. Threshold values of these two dynamic MR indexes were determined on the basis of positive differentiations.
RESULTS: The mean relative enhancement ratio and mean slope of enhancement for the malignant SPN group were significantly higher than those for the benign SPN group and significantly lower than those for the active infection group (P < .05). With 0.15 as the threshold maximum relative enhancement ratio for distinguishing the malignant SPN and active infection groups from the benign SPN group, the sensitivity, specificity, and accuracy were 100%, 70%, and 95%, respectively. With 0.025/sec as the threshold slope of enhancement, all SPNs with malignancy and active infection were clearly distinguished from benign SPNs.
CONCLUSION: Dynamic MR indexes were useful in the differentiation between SPNs that necessitated further evaluation or treatment (malignancy and active infection) and SPNs that did not necessitate further evaluation or treatment (benign nodules).
© RSNA, 2002
Index terms: Lung, diseases, 60.20, 60.21, 60.23, 60.31, 60.32, 60.33 • Lung neoplasms, 60.31, 60.32, 60.33 • Lung neoplasms, MR, 60.121412, 60.12143, 60.12144 • Lung nodule, 60.281, 60.30, 60.332 • Magnetic resonance (MR), perfusion study, 60.121412, 60.12143, 60.12144
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The blood supply and metabolism of malignant neoplasms are qualitatively and quantitatively different from those of chronic infection and benign neoplasms according to biologic and histopathologic findings (1–7,9–22). However, increased blood flow and capillary permeability are observed not only in malignant neoplasms but also in tissues with active infection, although the biochemical mechanisms and theories of these two diseases may be different. Moreover, increased FDG uptake due to increased expression of glucose transporter messenger RNA and the resulting increase in glucose transport protein in tissue with active infection has been observed; this phenomenon occurs similarly in malignant neoplasms (6,7). Therefore, accurately differentiating malignant SPNs from benign SPNs, especially those that coexist with active infection, on the basis of these biological and biochemical mechanisms may be difficult.
From a practical standpoint, it may be helpful when planning the management of SPNs to differentiate between the nodules that will necessitate further evaluation or treatment and those that will not necessitate further evaluation or treatment. Tissue perfusion and blood flow patterns in the liver (23), brain (24), and breast (25) have been investigated by using high temporal resolution and a sharp contrast material bolus profile at dynamic MR imaging. In addition, contrast material–enhanced dynamic MR imaging performed with an extremely short echo time has been shown to be useful for functional evaluation of the lung despite the very short T2* relaxation of the lung parenchyma (26–30).
In the present study, we hypothesized that contrast-enhanced dynamic MR imaging with extremely short repetition and echo times, owing to its high temporal resolution and sharp contrast material bolus profile, could enable the characterization of SPNs and might have a role in the management of SPNs by enabling differentiation between SPNs that necessitate further evaluation or treatment and those that do not necessitate further evaluation or treatment. Thus, the purpose of our study was to evaluate the utility of dynamic MR imaging in the management of SPNs.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Patients were selected for dynamic MR imaging according to the following criteria: (a) presence of SPN smaller than 30 mm in diameter with no calcifications, (b) absence of contraindications to the administration of gadolinium-based contrast material, and (c) ability to cooperate for the procedure. Nodule diameter was defined as the largest diameter on conventional radiographs and/or CT scans obtained with a lung window setting. "No calcification" meant SPNs with no visible calcification on conventional CT scans (ie, 1–10-mm section thickness) obtained with a mediastinal window setting. Fifty-eight patients (35 men, 23 women; mean age, 68 years; age range, 25–81 years) with 58 SPNs met the criteria and were referred for dynamic MR imaging. The final diagnoses were confirmed by means of microbiologic examination, cytologic examination, CT-guided transthoracic needle biopsy, transbronchial lung biopsy, video-assisted thoracoscopic surgery, or surgical resection, and pathologic examination in all cases. The institutional review board of Kobe University Graduate School of Medicine, Japan, approved the study, and informed consent was obtained from all patients before the MR studies were performed.
MR Imaging
All MR studies were performed by using a 1.5-T superconducting magnet (Gyroscan ACS-NT; Philips Medical Systems, Best, the Netherlands) and a body coil. Dynamic MR images (2.7/0.6 [repetition time msec/echo time msec], 20° flip angle, 128 x 96 matrix, 256 x 192 reconstructed matrix, 350–400 x 245–280-mm rectangular field of view) were acquired with a three-dimensional radio-frequency spoiled gradient-echo sequence. A three-dimensional slab thickness of 110 mm was used with 11 partitions by using overcontiguous sections in the coronal plane with a left-to-right phase-encoding direction; this protocol resulted in an effective partition thickness of 10 mm and real phase encoding of six steps in a section direction. The temporal resolution was 1.1 seconds for each three-dimensional data set. In all patients, a standard dose (0.1 mmol/kg body weight) of gadopentetate dimeglumine (Magnevist; Schering Japan, Osaka, Japan) was administered by means of an automatic infusion system (Sonic Shot; Nemoto, Tokyo, Japan) in a bolus through a cubital vein at a rate of 5 mL/sec. Twenty milliliters of saline solution was then administered at the same rate. The basic theory behind and application of contrast-enhanced dynamic MR imaging are documented in the literature (26,27).
At MR imaging, the patients inhaled 100% oxygen through a mask at a rate of 5 L/min. Before undergoing MR imaging, the patients received careful instruction in the breath-holding technique and practiced this to produce precisely the same degree of inspiration for each imaging series. At each acquisition, 30 images were obtained during 33 seconds of breath holding at end inspiration. All 58 MR perfusion imaging examinations were completed successfully. No adverse effects were observed.
Data and Statistical Analyses
Fifty-eight SPNs were classified into three groups on the basis of the final pathologic analysis–proved diagnosis: The malignant SPNs (n = 38) were 30 primary peripheral lung cancers (three localized bronchioloalveolar carcinomas, 19 adenocarcinomas, and eight squamous cell carcinomas) and eight metastatic lung tumors (six renal cell carcinomas and two thyroid cancers). The 10 benign SPNs were three hamartomas and seven tuberculomas. The 10 active infections were three active cases of tuberculosis, two atypical mycobacterial infections, one cryptococcosis infection, two aspergillosis infections, and two cases of organizing pneumonia. The tuberculomas were distinguished from active tuberculosis on the basis of the following criteria: (a) no evidence of change in size at follow-up CT performed every 6 months for more than 1 year and (b) no evidence of the presence of Mycobacterium tuberculosis at microbacterial examination.
One chest radiologist (Y.O.) with 8 years of experience drew regions of interest (diameter range, 8–29 mm) over the normal lung parenchyma and SPN on the MR images. The region of interest drawn over the tumor encompassed the entire cross-sectional area of the nodule and thus was as large as possible to minimize the effects of tumoral hemodynamic inhomogeneities. Signal intensity–time curves of tumor and normal lung parenchyma were created.
In this study, two indexes—maximum relative enhancement ratio and slope of enhancement—were calculated to characterize the SPNs. The maximum enhancement ratio (MER) was determined with the following equation: MER = (SImax - SI0)/SI0, where SImax is the maximum signal intensity after bolus injection of the contrast material and SI0 is the signal intensity at baseline. The maximum enhancement ratio of the tumor can be normalized by using the maximum enhancement ratio of the reference tissue (ie, normal lung parenchyma) to generate the maximum relative enhancement ratio.
The slope of enhancement (SLE, expressed in seconds-1) was determined with the following equation: SLE = maximum relative enhancement ratio/(tpeak - tstart), where tpeak is the time of maximum signal intensity after bolus injection of the contrast material and tstart is the time of the beginning of the continuous contrast enhancement effect of the SPN. The slope of enhancement of tumor represents the time of the first transit of the contrast material bolus (22).
Maximum relative enhancement ratios and slopes of enhancement were calculated by using the dynamic MR imaging data and compared among the three SPN groups by using analysis of variance. The Tukey honestly significant difference multiple comparison test was then performed. All values were expressed as means ± standard errors of the mean. A P value of less than .05 was considered to indicate a statistically significant difference.
Receiver operating characteristic analysis was performed to evaluate the usefulness of both the maximum relative enhancement ratio and the slope of enhancement as markers for distinguishing malignant SPNs from benign SPNs and for distinguishing malignant SPNs from active infections. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were calculated according to each value of these indexes and by varying the index values that indicated a positive differentiation (ie, threshold values) (17). In the differentiation between malignant and benign SPNs, sensitivity was defined as the percentage of malignant SPN cases that had an index greater than the given threshold value; specificity was defined as the percentage of benign SPN cases that had an index less than or equal to the threshold value. In the differentiation between malignant SPNs and active infections, sensitivity was defined as the percentage of malignant SPN cases that had an index lower than the threshold value; specificity was defined as the percentage of active infection cases that had an index greater than or equal to the given threshold value (2).
Feasible threshold dynamic MR indexes were tested for their capability to enable differentiation between malignant SPNs and active infection and between malignant SPNs and benign SPNs. Subsequently, these threshold values were tested for their capability to enable differentiation between the SPNs that necessitated further evaluation or treatment (ie, malignant SPN and active infection groups) and the SPNs that did not necessitate further evaluation or treatment (ie, benign SPN group).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The SPNs with active infection in this study showed complete overlap with the malignant SPNs when the maximum relative enhancement ratio and slope of enhancement were analyzed. The SPNs with active infection had even higher maximum relative enhancement ratios and slopes of relative enhancement compared with the malignant SPNs. However, an overlap of dynamic MR indexes between active infection and malignancy is expected when the underlying angiogenesis of a malignant neoplasm and inflammation is considered. It is well known that both malignant neoplasms and tissues with active infection have increased blood flow, perfusion, and capillary permeability at radiopathologic, pharmacokinetic, and pathologic studies (31–43).
Results of pharmacokinetic analysis have demonstrated that the increased signal intensity after contrast material injection depends on the perfusion, the relative blood volume, the capillary surface area, the permeability product, and the volume of extravascular fluid in the tissue. Initial changes in the signal intensity (or relative enhancement ratio)–time course curve at the time of the first transit of contrast material are considered to correlate with the combination of perfusion (blood flow per unit of tissue), increased extracellular space for accumulation of contrast material, and permeability of capillaries (31–43). All of these components have been reported to be markedly increased in malignant neoplasms.
Acute inflammation has three major components: (a) alterations in vascular caliber that lead to an increase in blood flow, (b) structural changes in the microvasculature that permit the plasma proteins and leukocytes to leave the circulation, and (c) leukocyte emigration from the microcirculation and accumulation in the focus of injury (42). These processes increase blood flow and vessel permeability and depend on the phase of inflammatory process. Although the underlying biological mechanisms of malignant neoplasms and active infections may be different, increased and rapid accumulation of contrast material was observed in both of these disease processes when dynamic MR indexes were analyzed.
Investigators have tried to differentiate malignant SPNs from benign SPNs previously (1–6,10–12). However, distinguishing active infection from a malignant neoplasm by using perfusion characteristics can be extremely difficult when the biological properties of malignant neoplasms and active infection that were just described are considered. Moreover, most of the cases of active infection that manifested as SPNs in our study necessitated intervention, such as bronchoalveolar lavage, transbronchial lung biopsy, CT-guided biopsy, or video-assisted thoracoscopic surgery, for determination of a final diagnosis. In the active infection group, the final diagnoses of only two cases (both active tuberculosis) were confirmed by means of microbiologic examinations of sputum. The final diagnoses of the other eight cases of active infection were confirmed as follows: bronchoalveolar lavage and/or transbronchial lung biopsy with microbiologic examination (n = 3: one case of active tuberculosis and two atypical mycobacterium infections), CT-guided transthoracic needle biopsy (n = 3: one cryptococcosis infection and two aspergillosis infections), video-assisted thoracoscopic surgery (one case of organizing pneumonia), and thoracotomy (one case of organizing pneumonia).
In eight (80%) of the 10 cases of active infection, an interventional examination was needed to differentiate the infections from the malignant tumors. With the exception of two cases of organizing pneumonia, all cases of active infection were subsequently treated medically. In our study, one case of organizing pneumonia was misdiagnosed as a malignancy at biopsy and the patient underwent video-assisted thoracoscopic surgery. In the other case of organizing pneumonia, the patient was highly suspected of having malignancy on the basis of cytologic specimen findings and therefore underwent thoracotomy and wedge resection.
In the management of SPNs, it may be helpful to distinguish the SPNs that necessitate further intervention or treatment from those that do not rather than differentiate between the malignant and benign SPNs. For this purpose, a dynamic MR imaging approach seems practical and reasonable. By using dynamic MR indexes, we were able to classify all the SPNs as either SPNs that necessitated subsequent evaluation or treatment or SPNs that did not necessitate subsequent evaluation or treatment.
Promising results in the differentiation between malignant and benign SPNs with dynamic CT and FDG PET have been reported previously (1–7). When Swenson et al (2) used dynamic CT and a threshold attenuation value of 20 HU to distinguish malignant SPNs from benign SPNs in 163 patients, they reported the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy to be 100%, 76.9%, 90.2%, 100%, and 92.6%, respectively. However, substantial overlap in enhancement among malignant neoplasms, granulomas, and benign neoplasms was reported (2). Similar overlap has been observed when only the maximum relative enhancement ratio at dynamic MR imaging was used. Yamashita et al (3) reported on 32 patients in whom dynamic CT enabled complete differentiation of lung cancers from tuberculomas and hamartomas when a maximum attenuation (ie, enhancement) of 20%–60% was adopted; however, cases of active infection were not included in this study group. Zhang and Kono (4) reported the possibility to more accurately differentiate malignant SPNs from benign SPNs in 65 patients by using the SPN-to-aorta ratio (parameter of SPN supplied from systemic circulation) with the peak height of enhancement. They also observed a substantial overlap in CT attenuation parameters among malignant, benign, and inflammatory nodules (4). The results of a more recent multicenter study (5) involving 356 patients demonstrated the sensitivity, specificity, and accuracy of CT for the detection of benignity to be 98%, 58%, and 77%, respectively, when a threshold attenuation value of 15 HU was used.
FDG PET reflects the glucose metabolism of various tissues. Patz et al (6) reported on 51 patients with SPNs for whom the sensitivity and specificity for detection of benign lesions were 89% and 100%, respectively, when a threshold standardized uptake ratio of 2.5 was used. Dewan et al (7), in a study involving 30 patients, reported that calculation of the differential uptake ratio enabled the differentiation between malignant and benign SPNs with a sensitivity of 95% and a specificity of 80%. Even with FDG PET, an overlap in differential uptake ratio between malignant SPNs and active Histoplasma infection was seen (7).
With use of the slope of enhancement, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy in distinguishing malignant SPNs from other SPN types were 100%, 85%, 93%, 100%, and 95%, respectively. Moreover, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of the slope of enhancement in the differentiation between SPNs that did and SPNs that did not necessitate further evaluation or treatment were 100%.
Our study results showed much less overlap compared with the overlap in the aforementioned studies because (a) we used dynamic MR indexes with a higher temporal resolution and a sharper contrast material bolus profile for evaluation of SPNs and (b) we treated the active infection group as part of the category of SPNs that necessitated further evaluation or treatment in our analyses. However, CT is a much more widely available modality, and FDG PET yields data on biological glucose metabolism, which may complement dynamic MR data in the evaluation of SPNs.
All of the SPNs included in this study were diagnosed by means of microbiologic and/or pathologic analysis. Therefore, there was no diagnostic dilemma in this study. The MR findings did not influence patient care in this study; however, they were observed to have a potential role in treatment management. A larger prospective comparative study with CT and PET, as well as a cost analysis, is needed to determine the true value of dynamic MR in the imaging of SPNs. A prospective study to compare dynamic CT, FDG PET, and dynamic MR indexes on the basis of our study results is planned for the near future.
There were some limitations in our study: (a) Although the described dynamic MR imaging technique involves the use of one of the fastest three-dimensional spoiled gradient-echo sequences currently available, even higher temporal resolution at dynamic MR imaging is desired, because the pulmonary circulation is 4.0–5.0 seconds and the pulmonary capillary circulation is 0.7 second in adults (43). (b) Ten metastatic lung tumors that were included in the malignant SPN group were highly vascular and could have biased the results. (c) The spatial resolution of the described dynamic MR imaging technique is lower than that achieved in prior studies. Therefore, we used dynamic MR data only for analysis of the SPN blood supply and not for morphologic analysis.
In conclusion, dynamic MR imaging enables accurate differentiation of primary and metastatic lung tumors from other SPN types. Moreover, SPNs can be successfully classified into two categories with use of dynamic MR imaging: SPNs that do and SPNs that do not necessitate further evaluation or treatment.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, Y.O.; study concepts and design, Y.O.; literature research, Y.O.; clinical studies, Y.O., K.S., S.A., D.T.; data acquisition, Y.O., D.T.; data analysis/interpretation, Y.O., H.H.; statistical analysis, Y.O.; manuscript preparation, Y.O., H.H.; manuscript definition of intellectual content, Y.O., H.H., D.T.; manuscript editing and revision/review, Y.O., H.H., M.K., K.S.; manuscript final version approval, Y.O., H.H.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Girvin and J. P. Ko Pulmonary Nodules: Detection, Assessment, and CAD Am. J. Roentgenol., October 1, 2008; 191(4): 1057 - 1069. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Zou, M. Zhang, Q. Wang, D. Shang, L. Wang, and G. Yu Quantitative Investigation of Solitary Pulmonary Nodules: Dynamic Contrast-Enhanced MRI and Histopathologic Analysis Am. J. Roentgenol., July 1, 2008; 191(1): 252 - 259. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A Yi, K. S. Lee, J. Han, M. P. Chung, M. J. Chung, and K. M. Shin 3-T MRI for Differentiating Inflammation- and Fibrosis-Predominant Lesions of Usual and Nonspecific Interstitial Pneumonia: Comparison Study with Pathologic Correlation Am. J. Roentgenol., April 1, 2008; 190(4): 878 - 885. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Cronin, B. A. Dwamena, A. M. Kelly, and R. C. Carlos Solitary Pulmonary Nodules: Meta-analytic Comparison of Cross-sectional Imaging Modalities for Diagnosis of Malignancy Radiology, March 1, 2008; 246(3): 772 - 782. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Y. Donmez, E. Yekeler, V. Saeidi, A. Tunaci, M. Tunaci, and G. Acunas Dynamic Contrast Enhancement Patterns of Solitary Pulmonary Nodules on 3D Gradient-Recalled Echo MRI Am. J. Roentgenol., December 1, 2007; 189(6): 1380 - 1386. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A Yi, T. Y. Jeon, K. S. Lee, J. H. Lee, J. B. Seo, Y. K. Kim, and M. J. Chung 3-T MRI: Usefulness for Evaluating Primary Lung Cancer and Small Nodules in Lobes Not Containing Primary Tumors Am. J. Roentgenol., August 1, 2007; 189(2): 386 - 392. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Kono, K. Fujimoto, H. Terasaki, N. L. Muller, S. Kato, J. Sadohara, N. Hayabuchi, and S. Takamori Dynamic MRI of Solitary Pulmonary Nodules: Comparison of Enhancement Patterns of Malignant and Benign Small Peripheral Lung Lesions Am. J. Roentgenol., January 1, 2007; 188(1): 26 - 36. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ohno, H. Hatabu, K. Murase, T. Higashino, M. Nogami, T. Yoshikawa, and K. Sugimura Primary Pulmonary Hypertension: 3D Dynamic Perfusion MRI for Quantitative Analysis of Regional Pulmonary Perfusion Am. J. Roentgenol., January 1, 2007; 188(1): 48 - 56. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. S. Shim, K. S. Lee, M. J. Chung, H. Kim, O J. Kwon, and S. Kim Do hemodynamic studies of stage t1 lung cancer enable the prediction of hilar or mediastinal nodal metastasis? Am. J. Roentgenol., April 1, 2006; 186(4): 981 - 988. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Schroeder, S. G. Ruehm, J. F. Debatin, M. E. Ladd, J. Barkhausen, and S. C. Goehde Detection of Pulmonary Nodules Using a 2D HASTE MR Sequence: Comparison with MDCT Am. J. Roentgenol., October 1, 2005; 185(4): 979 - 984. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. K. Shankar and D. C. Sullivan Functional Imaging in Lung Cancer J. Clin. Oncol., May 10, 2005; 23(14): 3203 - 3211. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A Yi, K. S. Lee, E. A Kim, J. Han, H. Kim, O J. Kwon, Y. J. Jeong, and S. Kim Solitary Pulmonary Nodules: Dynamic Enhanced Multi-Detector Row CT Study and Comparison with Vascular Endothelial Growth Factor and Microvessel Density Radiology, October 1, 2004; 233(1): 191 - 199. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Schaefer, J. Vollmar, F. Schick, R. Vonthein, M. D. Seemann, H. Aebert, R. Dierkesmann, G. Friedel, and C. D. Claussen Solitary Pulmonary Nodules: Dynamic Contrast-enhanced MR Imaging--Perfusion Differences in Malignant and Benign Lesions Radiology, August 1, 2004; 232(2): 544 - 553. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. M. Vogt, C. U. Herborn, P. Hunold, T. C. Lauenstein, T. Schroder, J. F. Debatin, and J. Barkhausen HASTE MRI Versus Chest Radiography in the Detection of Pulmonary Nodules: Comparison with MDCT Am. J. Roentgenol., July 1, 2004; 183(1): 71 - 78. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||
= sensitivity for distinguishing malignant SPNs from benign SPNs group, 
= specificity for distinguishing malignant SPNs from benign SPNs, 
= sensitivity for distinguishing malignant SPNs group from active infections, 
= specificity for distinguishing malignant SPNs from active infections.
