| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ultrasonography |
1 From the Dept of Imaging, Hammersmith Hospital, Imperial College School of Medicine, 150 du Cane Rd, London W12 0HS, England (M.J.K.B., D.O.C., C.J.H., R.A.H., J.B.B., R.J.E.); Dept of Radiology, Kings College Hospital, London, England (P.S.S.); Dept of Radiology, Klinikum Benjamin Franklin, Berlin, Germany (T.A.); and Dept of Radiology, Univ of Chieti, Italy (R.B.). From the 1999 RSNA scientific assembly. Received Nov 23, 1999; revision requested Jan 5, 2000; final revision received Mar 6, 2001; accepted Apr 6. M.J.K.B. and C.J.H. supported by the Medical Research Council of the United Kingdom. Supported by Schering, Berlin, Germany, and Acuson, Stockley Park, England. Address correspondence to M.J.K.B. (e-mail: m.blomley@ic.ac.uk).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Thirty-seven patients with characterized lesions (metastasis, n = 17; hepatocellular carcinoma, n = 4; hemangioma, n = 9; focal nodular hyperplasia, n = 7) received 2.5 g SH U 508A. After 5 minutes, stimulated acoustic emission was elicited by using a previously described method. Liver and/or lesional differences were assessed with videodensitometry (objective conspicuity score), and two observers assessed each lesion by using a six-point scale (subjective conspicuity score).
RESULTS: Metastases and hepatocellular carcinoma had low stimulated acoustic emission; median objective conspicuity scores were 70% and 68% (all scores were 
43%), respectively, and subjective conspicuity scores were 2 or higher for both observers. Hemangiomas had reduced stimulated acoustic emission, with more variability; the median objective conspicuity score was 41% (range, 9%–72%), and the median subjective conspicuity scores were 2 (range, 1–4) and 3.5 (range, 1–5) for observers 1 and 2, respectively. Focal nodular hyperplasia had stimulated acoustic emission comparable to that of the liver in all cases; the median objective conspicuity score was -4.7% (all scores were <6%), and the subjective conspicuity score was 1 or lower for both observers. This finding completely separated focal nodular hyperplasia and malignancies. Significant differences were seen between focal nodular hyperplasia and all other lesion types (P < .05).
CONCLUSION: Strong late-phase lesional uptake of SH U 508A is characteristic of focal nodular hyperplasia, is seen in some hemangiomas, and was not observed in malignancies.
Index terms: Liver neoplasms, diagnosis, 761.3198, 761.3194, 761.323, 761.33 • Liver neoplasms, US, 761.1296, 761.12981, 761.12983, 761.12988 • Ultrasound (US), contrast media, 761.12988
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
SH U 508A consists of galactose granules and water, with a small admixture of palmitic acid (0.1%). When reconstituted, multiple small, stabilized air microbubbles are produced. SH U 508A produces Doppler enhancement of the blood pool for about 2–5 minutes (6), with an additional delayed liver and splenic phase lasting longer than 30 minutes (7). Imaging of the liver phase of SH U 508A requires the use of special scanning techniques, such as microbubble disruption with color Doppler techniques (stimulated acoustic emission [SAE]) (5,8) or harmonic methods, such as pulse inversion imaging (9) and intermittent second-harmonic US (10). Recent studies (5,11) have demonstrated that the conspicuity of liver metastases can be increased, because they show reduced signal on a background of transient parenchymal enhancement. SAE uses color Doppler scanning to disrupt the microbubbles; their sudden disappearance is interpreted in the autocorrelation process as strong transient color enhancement. With SAE, microbubble-specific and conventional images can be simultaneously captured by switching on and off the color overlay on a frozen image.
Previous investigators used these methods to improve the sensitivity of liver US. On the basis of data obtained by using liver-specific agents in nuclear medicine and magnetic resonance (MR) imaging, we hypothesized that specificity could also be improved. High uptake of sulphur colloid or superparamagnetic iron oxide agents is known to be a feature of focal nodular hyperplasia (FNH) (12,13) and can reveal their characteristic central scars (14). This uptake is attributable to the presence of Kupffer cells within these lesions. Uptake has also been described in other types of benign liver tumors, including hemangiomas, adenomas, and regenerating nodules. Proposed mechanisms include reticuloendothelial uptake and delayed blood pool enhancement. By contrast, malignant lesions, such as metastases and hepatocellular carcinomas (HCCs), usually show little uptake, although well-differentiated HCCs can be an exception (15). By analogy, we hypothesized that liver-specific microbubble agents have similar effects. We therefore designed a study to compare the amount of uptake of SH U 508A in different types of liver lesions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Seventeen (five men, 12 women; mean age, 62 years; age range, 37–77 years) of the 37 had liver metastases. In nine, the metastases were proven at liver biopsy, while in eight, the diagnosis was based on previous reports of liver metastases from both computed tomographic (CT) and US examinations performed in patients with a known malignancy and evidence of lesion enlargement from repeat imaging studies performed during less than a year. In 10 patients, the primary tumor was of a type usually classified as hypovascular (pancreatic adenocarcinoma, n = 3; colorectal adenocarcinoma, n = 2; undifferentiated adenocarcinoma, n = 2; lung, ovary, or gastric carcinoma, n = 1 each), while in seven, it was hypervascular (carcinoid, n = 5; renal carcinoma or sarcoma, n = 1 each).
Four (three men, one woman; mean age, 63 years; age range, 54–77 years) of the 37 patients had biopsy-proven HCC. Three had moderately differentiated HCC that had arisen on a background of liver cirrhosis. In one, a well-differentiated HCC was found with no evidence of liver cirrhosis.
Nine (six men, three women; mean age, 47 years; age range, 31–66 years) patients had hemangiomas. The diagnosis was made at histologic examination in three, while in six, characteristic features of hemangiomas had been demonstrated at both contrast material–enhanced CT and US. The CT scans were obtained by using a dynamic serial hemangioma protocol (17) with complete or near-complete infilling of the lesion with contrast material on delayed scans obtained 5–10 minutes after the injection of contrast material. All six patients had stable disease with no evidence of lesion enlargement at sequential CT and/or US examinations performed at least 1 year apart.
Seven (one man, six women; mean age, 45 years; age range, 34–60 years) of the 37 patients had FNH. In five, FNH was proven at biopsy. In two, previous US images, CT images, and angiograms (all of which had been obtained at least once in each patient) had been reported to be characteristic of FNH. Both of these patients were also known to have no evidence of enlargement at sequential CT and US examinations performed at least 2 years apart.
In addition to these 37 patients with definitively characterized lesions, five patients were recruited for this study, and their data were analyzed. However, these patients were subsequently excluded from the final analysis because the diagnosis was not convincingly proven. The clinical diagnoses in these patients were multiple adenomas, FNH (n = 2), HCC, and hemangioma. The details were as follows: (a) Adenoma was presumed in a male patient, aged 30 years, with multiple vascular liver lesions that complicated longstanding glycogen storage disease, which remained unchanged on sequential CT and US scans obtained during 3 years. (b) FNH was presumed in a female patient, aged 44 years, with a hypervascular liver lesion at US and CT that had been stable during 4 years. (c) HCC was presumed in a male patient, aged 63 years, with cirrhosis and multiple focal liver lesions that appeared to be consistent with HCC at angiography and lipiodol CT. (d) FNH was presumed in an otherwise healthy female patient, aged 34 years, with a hypervascular liver lesion at US and CT. (e) Hemangioma was presumed in a female patient, aged 64 years, with esophageal carcinoma and a well-defined echogenic liver lesion that showed features characteristic of hemangioma on a tailored CT scan. For completeness, data from these five patients are listed in the Results section but were not included in any statistical analyses.
In each patient, a bolus of 2.5 g SH U 508A with a concentration of 300 mg/mL was intravenously injected; this injection was followed with a 10-mL normal saline flush. After a delay of 5 minutes to prevent the destruction of any microbubbles that might have accumulated in the liver, the lesion was scanned by using a method previously described (5) to elicit stimulated acoustic emission. Briefly, the lesion was localized with gray-scale imaging, and color Doppler imaging was then activated by using a color box that enclosed the lesion. We used machine settings that were known to work well in eliciting SAE: a Doppler frequency of 2.5 MHz, maximum acoustic power, and pulse repetition frequency settings (5). The same scanner and probe were used throughout (Sequoia 512, 4V2 probe; Acuson, Mountain View, Calif). While scanning a lesion, we moved the focal zone and probe positions so that SAE (which was known, based on previous findings, to be dependent on these settings) could be elicited to the maximal extent possible throughout the lesion and adjacent liver. Several sonologists (M.J.K.B., P.S.S., D.O.C.) performed the scanning. The studies were recorded on S-VHS videotape, and a representative segment was chosen in each patient. This segment was used for both objective and subjective analyses.
The objective analysis was performed by using commercial videodensitometric software (CQ; Kinetic Imaging, Liverpool, England) and a personal computer equipped with a frame-grabbing card (Meteor; Matrox Graphics, Dorval, Quebec, Canada). One of the authors (M.J.K.B.) performed all of the videodensitometric analyses. A video sequence was grabbed and reviewed. The color data were segmented from the B-mode image by using the software. The image in the sequence that showed maximal SAE in the lesion and adjacent liver was selected for further analysis.
Regions of interest were drawn in the areas of maximal SAE within the lesion and in the adjacent liver at the same depth. The regions of interest (drawn with a minimum size of 100 pixels) were located clearly within the lesion or liver, and were not on the border between the two to avoid potential saturation effects of the SAE signal. The regions of interest ranged from 109 to 2,911 pixels (mean, 1,036 pixels) in the liver and from 159 to 7,560 pixels (mean, 921 pixels) in the lesions. The color data were automatically segmented from the B-mode image by using the software. The proportion of pixels that were colored was expressed as a percentage of the region of interest, and the difference between the liver and region-of-interest scores was calculated as an objective conspicuity score (OCS). For example, if regions of interest of 1,000 pixels were drawn in the liver and lesion and if the region for the liver had 900 color pixels and that for the lesion had 100, the proportionate score for the liver would be 90%, while the proportionate score for the lesion would be 10%. The OCS would then be 90% – 10%, or 80%.
The same videotaped sequences were also presented to two independent observers (T.A., C.J.H.). Thus, the observers were given only the video sequences that showed the SAE data. They were blinded to clinical and other imaging data and to data from other parts of the US study, such as conventional Doppler data. However, it was impossible to completely blind them to the conventional B-mode appearances. They scored the amount of color SAE within the lesions and adjacent liver on the videotaped data by using the following six-point scale: 0, no effect; 1, as many as 25% color pixels at the level of the focal zone; 2, 26%–50% color pixels at the level of the focal zone; 3, 51%–75% color pixels at the level of the focal zone; 4, 76%–100% color pixels at the level of the focal zone; and 5, 76%-100% color pixels for 2 cm on either side of the level of the focal zone.
As described previously, the sequences were dynamic, and the probe and focal zone positions were adjusted so that SAE was elicited in different areas. When different parts of a lesion or adjacent liver showed different amounts of enhancement, the observers were asked to score the area showing maximal enhancement. The observers were asked to analyze areas of liver that were as close as possible to the lesion. The reason for this approach was that previous findings (7) showed that SAE is a depth-dependent effect.
As with the computer-aided analysis, the relative amount of liver enhancement was normalized. This normalization was performed by subtracting the differences between the SAE scores within and outside the lesion to determine the subjective conspicuity score (SCS). For example, if the lesion had a score of 3 and the adjacent liver had a score of 5, the SCS was 2.
The OCSs and SCSs were then analyzed by using nonparametric methods with the program INSTAT (Graph Pad Software, San Diego, Calif) in consultation with a professional medical statistician. All comparisons were performed independently and separately for the OCSs and each observer’s SCSs. In the first part of the analysis, the four categories of lesion type (ie, metastases, HCC, hemangioma, and FNH) were separately analyzed to determine if the lesions showed reduced SAE compared with that of the adjacent liver. A nonparametric two-tailed Wilcoxon signed rank test was performed to test the null hypotheses that the median OCSs and SCSs were zero. In the second part of the analysis, groups were compared by using an unpaired nonparametric Kruskal-Wallis test of all lesion types to determine if they differed. The data were further analyzed by using a Wilcoxon two-sample rank sum comparison with Bonferroni correction. In addition, both unweighted and weighted 
statistics were calculated for the subjective analyses; the weighting was proportional to the distance from the diagonal line that indicated agreement (18).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The median, minimum, and maximum OCSs and SCSs for each diagnostic group are shown for both observers (Fig 1; Table).
|
|
|
|
|
|
|
|
|
|
statistics for interobserver variation were 0.21 and 0.51, respectively. Although not included in the previous statistical analysis, data from the five excluded patients supported the findings in the rest of the study. The case of presumptive HCC showed a large difference between the lesion and the liver (OCS, 79%; SCS for observer 1, 4; SCS for observer 2, 2), while both cases of presumptive FNH showed small differences (OCSs 11% and 25%; SCSs for observer 1, 0 and 0; SCSs for observer 2, 0 and 1). The presumptive hemangioma also showed high SAE (OCS, 9%; SCS for observer 1, 1; SCS for observer 2, 0). The presumed adenoma showed low signal (OCS 80%; SCS for observer 1, 3; SCS for observer 2, 5).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To our knowledge, these observations have not been previously described with US, but our findings are analogous to those observed at MR imaging with superparamagnetic iron oxide agents, if imaging is performed during the liver-specific phase of these agents. In general, benign lesions show enhancement, while malignancies do not (13,14). Vogl et al (16) studied the superparamagnetic iron oxide agent AMI-25 and found that normal liver, FNH, and adenomas all showed enhancement, while metastases and HCC did not. The central scar of FNH was sometimes better shown. Similar findings were reported by Denys et al (19), who, interestingly, also observed that hemangiomas showed an intermediate degree of enhancement. In a separate study with another superparamagnetic iron oxide agent, SH U 555A, Vogl et al (20) had similar findings, which included no signal loss in metastases and HCC, moderate loss in hemangiomas, and marked effects in FNH. These differences are thought to be due to the presence of Kupffer cells in FNH, although the reason for the enhancement changes in hemangiomas is not clear, because they are not thought to contain Kupffer cells. Delayed clearance of contrast material from the blood pool, which is comparable to the accumulation of labeled red cells, is a more likely mechanism (13,19).
Probably, similar mechanisms are at work with SH U 508A, and the striking analogy of our data to those obtained with superparamagnetic iron oxide agents suggests that the kinetics of the two types of contrast agent are similar. The liver phase of at least one other microbubble agent is mediated by Kupffer cell uptake (21), although the mechanism by which this occurs with SH U 508A is not yet understood. Interaction with Kupffer cells would explain the enhancement that we observed in FNH at 5 minutes after injection. In our series, the enhancement seen in some hemangiomas is presumably attributable to blood pooling, with stationary or slowly moving microbubbles collecting in the vascular spaces, which are features of these lesions. While the duration of useful enhancement in the main vessels is only 2–3 minutes, separate work (7) has shown that some vascular-phase effects in the kidney can last for 5 minutes and longer. Unfortunately, it would be difficult to perform sequential dynamic studies to investigate this finding, because each time scanning is performed, the ultrasound beam disrupts some or all of the microbubbles in a region of interest.
There are four main limitations of this study. The first is the small number of patients, especially those with HCC. The investigation of possible differences between different types of HCC requires a larger series. Well-differentiated HCCs show more uptake of superparamagnetic iron oxide agents compared with less-differentiated HCCs because they may contain a greater proportion of Kupffer cells (13). Therefore, overlap is possible between well-differentiated HCCs, for example, and other types of liver lesions, such as FNH. However, the fact that we did not observe this overlap in our small sample of four patients is noteworthy. Indeed, both observers recorded no SAE at all within the one well-differentiated HCC studied. The inclusion of patients with other conditions, such as macroregenerative nodules, adenoma, and focal fatty change or sparing, would have been helpful in the investigation of late-phase uptake in these lesions, because all of these lesions should contain Kupffer cells. Although not acceptable for the formal statistical analysis, the data from the patients with incompletely characterized lesions completely agreed with that of the rest of the study; this agreement further supported our findings.
A second limitation is the lack of biopsy correlation in some patients. However, in the 37 patients included in this statistical analysis, lesions that were not examined at biopsy were well characterized at multiple sequential imaging examinations during at least 1 year. Furthermore, the same separation and statistical differences were seen between FNH and other lesions when the two patients with FNH who had not undergone biopsy were excluded from the statistical analysis.
A third limitation is the method of assessment, both subjective and objective. Computer-aided videodensitometry, as performed here, is subject to a number of errors, including the potential for bias in selecting a particular frame for analysis and the fact that no direct relationship exists between the number of color pixels on a videotaped image obtained at Doppler US based on mean frequency and the number of microbubbles. It does, however, add a measure of objectivity to the subjective analyses, which can be influenced by other factors (such as the conventional B-mode appearances of a lesion). Subjective assessments are obviously subject to perceptual differences, but they allow easy assessment of dynamic sequences that include changes in scanning plane, probe position, and focal zone. In the present study, these changes were often needed to elicit the maximal SAE effect in each area. While the scores of the two observers varied, we are encouraged that conclusions from the data for both observers and from the objective analysis were essentially identical. This finding suggests that this technique can be used for the subjective analysis of a lesion, without offline analyses.
A fourth limitation is that the scanning time, 5 minutes after injection, was empirical; in a previous study (7), this time produced good liver-specific enhancement with SH U 508A. Different results may have been seen with later or earlier scanning. For example, more enhancement might have been seen in hemangiomas if SAE had been elicited at 3 minutes, because the blood pool effect may be stronger at this time. Unfortunately, studies to evaluate the optimal scanning time are difficult to perform because the act of scanning inactivates the contrast agent; therefore, sequential scanning provides a false impression of the kinetics (5,7). These issues should be addressed in future work.
In summary, the strong SAE that we found in FNH clearly separates FNH from other lesions, and we propose that late-phase scanning with SH U 508A may be useful as a noninvasive confirmatory test for this condition. This test would be particularly useful because FNH is often detected at US but then requires other confirmatory tests and often biopsy for confident characterization. These findings may also tentatively suggest that late-phase SAE at imaging may be useful in characterizing other lesions, such as hemangioma, in which we observed more late-phase activity compared with that of malignant lesions. By varying the times at which SAE is elicited, a clearer delineation may be observed. Larger multicenter studies will be helpful in evaluating these points and in investigating the reproducibility of this method.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, M.J.K.B.; study concepts, M.J.K.B.; study design, M.J.K.B., D.O.C., T.A., P.S.S., R.J.E.; literature research, M.J.K.B., R.B.; clinical studies, M.J.K.B., P.S.S., D.O.C., R.A.H., R.B., J.B.B..; data acquisition, M.J.K.B., P.S.S., D.O.C., R.A.H., R.B., J.B.B.; data analysis, M.J.K.B., T.A., C.J.H., R.J.E.; statistical analysis, M.J.K.B.; manuscript preparation, M.J.K.B.; manuscript definition of intellectual content, M.J.K.B., D.O.C.; manuscript editing, review, and final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. Quaia, A. Palumbo, S. Rossi, F. Degobbis, S. Cernic, G. Tona, and M. Cova Comparison of visual and quantitative analysis for characterization of insonated liver tumors after microbubble contrast injection. Am. J. Roentgenol., June 1, 2006; 186(6): 1560 - 1570. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V R Stewart and P S Sidhu New directions in ultrasound: microbubble contrast. Br. J. Radiol., March 1, 2006; 79(939): 188 - 194. [Full Text] [PDF]
|
|
|
|
|
|
C. Nicolau, R. Vilana, V. Catala, L. Bianchi, R. Gilabert, A. Garcia, and C. Bru Importance of Evaluating All Vascular Phases on Contrast-Enhanced Sonography in the Differentiation of Benign from Malignant Focal Liver Lesions Am. J. Roentgenol., January 1, 2006; 186(1): 158 - 167. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Takeshima, T. Kumada, H. Toyoda, S. Kiriyama, M. Tanikawa, H. Ichikawa, T. Kawachi, and S. Ogawa Comparison of IV Contrast-Enhanced Sonography and Histopathology of Pancreatic Cancer Am. J. Roentgenol., November 1, 2005; 185(5): 1193 - 1200. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Celik, H. Ozdemir, C. Yucel, S. Gultekin, S. O. Oktar, and M. Arac Characterization of Hyperechoic Focal Liver Lesions: Quantitative Evaluation With Pulse Inversion Harmonic Imaging in the Late Phase of Levovist J. Ultrasound Med., January 1, 2005; 24(1): 39 - 47. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. von Herbay, C. Vogt, R. Willers, and D. Haussinger Real-time Imaging With the Sonographic Contrast Agent SonoVue: Differentiation Between Benign and Malignant Hepatic Lesions J. Ultrasound Med., December 1, 2004; 23(12): 1557 - 1568. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. H. Bryant, M. J. Blomley, T. Albrecht, P. S. Sidhu, E. L. S. Leen, R. Basilico, J. M. Pilcher, L. H. Bushby, C. W. Hoffmann, C. J. Harvey, et al. Improved Characterization of Liver Lesions with Liver-Phase Uptake of Liver-specific Microbubbles: Prospective Multicenter Study Radiology, September 1, 2004; 232(3): 799 - 809. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Quaia, F. Calliada, M. Bertolotto, S. Rossi, L. Garioni, L. Rosa, and R. Pozzi-Mucelli Characterization of Focal Liver Lesions with Contrast-specific US Modes and a Sulfur Hexafluoride-filled Microbubble Contrast Agent: Diagnostic Performance and Confidence Radiology, August 1, 2004; 232(2): 420 - 430. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Migaleddu, G. Virgilio, D. Turilli, M. Conti, G. Campisi, N. Canu, D. Sirigu, and I. Vincentelli Characterization of Focal Liver Lesions in Real Time Using Harmonic Imaging with High Mechanical Index and Contrast Agent Levovist Am. J. Roentgenol., June 1, 2004; 182(6): 1505 - 1512. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C F Dietrich, A Ignee, J Trojan, C Fellbaum, and G Schuessler Improved characterisation of histologically proven liver tumours by contrast enhanced ultrasonography during the portal venous and specific late phase of SHU 508A Gut, March 1, 2004; 53(3): 401 - 405. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. von Herbay, C. Vogt, and D. Haussinger Differentiation Between Benign and Malignant Hepatic Lesions: Utility of Color Stimulated Acoustic Emission With the Microbubble Contrast Agent Levovist J. Ultrasound Med., February 1, 2004; 23(2): 207 - 215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H-D Liang and M J K Blomley The role of ultrasound in molecular imaging Br. J. Radiol., December 1, 2003; 76(suppl_2): S140 - S150. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Youk, C. S. Kim, and J. M. Lee Contrast-Enhanced Agent Detection Imaging: Value in the Characterization of Focal Hepatic Lesions J. Ultrasound Med., September 1, 2003; 22(9): 897 - 910. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Youk, J. M. Lee, and C. S. Kim Therapeutic Response Evaluation of Malignant Hepatic Masses Treated by Interventional Procedures With Contrast-Enhanced Agent Detection Imaging J. Ultrasound Med., September 1, 2003; 22(9): 911 - 920. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Albrecht, M. J. K. Blomley, P. N. Burns, S. Wilson, C. J. Harvey, E. Leen, M. Claudon, F. Calliada, J.-M. Correas, M. LaFortune, et al. Improved Detection of Hepatic Metastases with Pulse-Inversion US during the Liver-specific Phase of SHU 508A: Multicenter Study Radiology, May 1, 2003; 227(2): 361 - 370. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A K P Lim, N Patel, W M W Gedroyc, M J K Blomley, G Hamilton, and S D Taylor-Robinson Hepatocellular adenoma: diagnostic difficulties and novel imaging techniques Br. J. Radiol., August 1, 2002; 75(896): 695 - 699. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS |
