DOI: 10.1148/radiol.2322031401
(Radiology 2004;232:420-430.)
© RSNA, 2004
Characterization of Focal Liver Lesions with Contrast-specific US Modes and a Sulfur Hexafluoride–filled Microbubble Contrast Agent: Diagnostic Performance and Confidence1
Emilio Quaia, MD,
Fabrizio Calliada, MD,
Michele Bertolotto, MD,
Sandro Rossi, MD,
Lorena Garioni, MD,
Laura Rosa, MD and
Roberto Pozzi-Mucelli, MD
1 From the Department of Radiology, Cattinara Hospital, University of Trieste, Strada di Fiume 447, Trieste 34149, Italy (E.Q., M.B., R.P.M.); and Department of Radiology (F.C., L.G.) and Operative Unit of Interventional Ultrasound (S.R., L.R.), San Matteo Hospital, University of Pavia, Italy. From the 2003 RSNA scientific assembly. Received August 31, 2003; revision requested November 11; revision received November 17; accepted January 15, 2004. Address correspondence to E.Q. (e-mail: equaia@yahoo.com).
 |
ABSTRACT
|
|---|
PURPOSE: To assess whether characterization of solid focal liver lesions could be improved by using ultrasonographic (US) contrast-specific modes after sulfur hexafluoride–filled microbubble contrast agent injection, as compared with lesion characterization achieved with preliminary baseline US.
MATERIALS AND METHODS: Four hundred fifty-two solid focal hepatic lesions that were considered indeterminate at baseline gray-scale and color Doppler US were examined after microbubble contrast agent injection performed by using low-acoustic-power contrast-specific modes during the arterial (10–40 seconds after injection), portal venous (50–90 seconds after injection), and late (100–300 seconds after injection) phases. Two readers independently and retrospectively reviewed baseline and contrast material–enhanced US scans and classified each depicted lesion as malignant or benign according to standard diagnostic criteria. Sensitivity, specificity, accuracy, and positive and negative predictive values and areas under the receiver operating characteristic curve (Az) were calculated by considering histologic analysis (317 patients) or contrast-enhanced helical computed tomography followed by serial US 3–6 months apart (135 patients) as the reference standards.
RESULTS: Different contrast enhancement patterns were observed according to lesion characteristics. During the late phase, benign lesions were predominantly hyper- or isoechoic relative to the adjacent liver parenchyma, whereas malignant lesions were predominantly hypoechoic. Review of the contrast-enhanced US scans after baseline image review yielded significantly improved diagnostic performance (P < .05). Overall diagnostic accuracy was 49% before versus 85% after review of the contrast-enhanced scan for reader 1 and 51% before versus 88% after review of the contrast-enhanced scan for reader 2. Diagnostic confidence—that is, the Az—was 0.820 before versus 0.968 after review of the contrast-enhanced scan for reader 1 and 0.831 before versus 0.978 after review of the contrast-enhanced scan for reader 2.
CONCLUSION: The use of contrast-specific modes with a sulfur hexafluoride contrast agent led to improved characterization of solid focal liver lesions.
© RSNA, 2004
Index terms: Liver neoplasms, 761.3192, 761.3194, 761.3198, 761.321, 761.323, 761.33 • Liver neoplasms, diagnosis, 761.30 • Liver neoplasms, US, 761.12981, 761.12983, 761.12984, 761.12988, 761.12989 • Microbubbles • Ultrasound (US), contrast media, 761.12988
|
INTRODUCTION
|
|---|
Baseline gray-scale and color Doppler ultrasonographic (US) examinations have limited accuracy in the characterization of solid focal liver lesions (1,2) because the depicted benign and malignant lesions may have similar echo patterns and vascular architectures. Although tumoral vessel visibility may be improved (3) by using color and power Doppler US after microbubble-based contrast agent injection, color signal saturation and blooming artifacts (4) represent important limitations.
Specialized contrast-specific US modes that enable one to overcome the limitations of baseline gray-scale and color Doppler US have been introduced and have been shown to lead to improved diagnostic performance in the characterization of focal liver lesions after the injection of SH U 508A (Levovist; Schering, Berlin, Germany), an air-filled microbubble contrast agent (5–10). SH U 508A has to be insonated by using high acoustic power to produce bubble destruction, with emission of a wideband frequency signal that is detectable by using contrast-specific modes. Although destructive US enhanced with SH U 508A requires intermittent scanning and a limited number of insonations to minimize bubble rupture, bubble destruction does occur, and, thus, prolonged evaluation of liver contrast enhancement cannot be performed. Now, however, these limitations can be overcome by performing nondestructive low-acoustic-power US scanning (11) with perfluorocarbon or sulfur hexafluoride–filled microbubbles (12–15).
SonoVue (BR1; Bracco, Milan, Italy) is a sulfur hexafluoride–filled microbubble contrast agent that is licensed for use in abdominal and vascular imaging in most European countries. This agent has a strong nonlinear harmonic response when it is insonated with low acoustic power (11–15). The safety and effectiveness of this agent in vascular and parenchymal diagnostic applications have been proved in preliminary experimental and clinical investigations (16). However, to our knowledge, the diagnostic performance of low-acoustic-power US enhanced with sulfur hexafluoride–filled microbubbles in the characterization of focal liver lesions has not yet been investigated in a large series of patients.
The aim of this study was to assess whether the characterization of solid focal liver lesions could be improved by using contrast-specific US modes after sulfur hexafluoride–filled microbubble injection, as compared with the lesion characterization achieved by using preliminary baseline US.
|
MATERIALS AND METHODS
|
|---|
Study Population
The study population consisted of 452 consecutive patients (mean age, 64 years ± 12 [standard deviation]; median age, 67 years; age range, 16–87 years) with solid focal liver lesions that were identified at baseline gray-scale US during routine clinical work-up and were considered indeterminate by the on-site sonologist. These patients—253 men (mean age, 65 years ± 11; median age, 68 years; age range, 16–83 years) and 199 women (mean age, 62 years ± 14; median age, 65 years; age range, 28–87 years)—were recruited from Cattinara Hospital and San Matteo Hospital during an 18-month period. The difference in median age between the male and female patients was found to be statistically significant (P < .05, Nonparametric Mann-Whitney U test performed after Shapiro-Wilk test results failed to show a normal distribution for age data). In 47 patients, who had more than one lesion with a similar appearance at baseline gray-scale US, the largest and most conspicuous lesion was selected for evaluation.
Patients who were critically ill, were hypersensitive to drugs, had severe heart disease, and/or had lesions that had been characterized at baseline US (ie, typical hemangiomas in patients without cirrhosis or in patients with no history of malignancy, obvious cysts, some focal fatty sparings, and multiple [>four] metastases) were excluded from the study. Informed consent was obtained from all patients after the nature of the procedure had been fully explained. The ethical principles of the Declaration of Helsinki (17) were strictly followed. No approval was required by the ethics review board of either of the two hospitals involved in this clinical study because the described sulfur hexafluoride contrast agent is licensed for use in liver imaging in our country. In addition, neither ethics review board approval nor informed consent was required for blinded retrospective off-site analysis.
Sonologists
Four sonologists, two from each hospital (E.Q. and M.B. from Cattinara Hospital; F.C. and S.R. from San Matteo Hospital), who had 8 (E.Q.), 11 (M.B.), 15 (F.C.), and 16 (S.R.) years of experience in liver US imaging were involved in this study. Each on-site sonologist had at least 5 years of experience in microbubble contrast material–enhanced US of the liver, was aware of the patients’ clinical histories, and was blinded to the biopsy results and all imaging findings except those of US.
Baseline US
Different US systems were used (Table 1). First, each lesion was scanned at gray-scale US by using tissue harmonic and compound US imaging examinations, which reduce noise and speckles. Each focal liver lesion was measured and assigned a liver segment location according to Couinaud (18) and Bismuth (19) classification systems. Tumoral vessels were imaged at color Doppler US by using slow-flow settings (ie, pulse repetition frequencies of 800–1,500 Hz, wall filters of 40–50 Hz, and high levels of color vs echo priority and color persistence) for optimal visualization of slow blood flows. Doppler US spectral analysis of tumoral vessels was performed to identify pulsatile arterial or continuous venous flow.
Contrast-enhanced US
After baseline US scan evaluation, the sonologist initiated the contrast-specific mode (Table 1). Before microbubble contrast agent injection, the persistence of the image display on the US machine was set to zero, the signal gain was registered below the noise threshold, and one focus was positioned below the level of the lesion. The microbubble contrast agent (sulfur hexafluoride) was injected intravenously as a bolus of 2.4 mL (5 mg/mL) by using a 20-gauge cannula; a 10-mL normal saline flush followed. In 90 of the 452 patients, contrast enhancement was considered inadequate by the on-site sonologist after the first bolus injection; these patients received an additional microbubble contrast agent bolus of 4.8 mL. Digital cine clips were stored on a personal computer (Intel, Pentium 4; Intel, Santa Clara, Calif) that was connected to the US equipment by way of a high-performance, hardware-based, real-time Moving Picture Experts Group 2 encoder (Mediacruise VR1000; Canopus, San Jose, Calif) and frame-grabber software (Mediacruise). Digital cine clips were registered during baseline US scanning and during the arterial (ie, 10–40 seconds from beginning of contrast agent bolus injection), portal venous (ie, 50–90 seconds from beginning of injection), and late (ie, 100–300 seconds from beginning of injection) phases.
On-Site Analysis
Immediately after the contrast-enhanced US scans were obtained, lesion contrast enhancement was subjectively assessed in consensus between the two sonologists at each hospital (E.Q. and M.B., F.C. and S.R.). Lesions with predominantly higher, similar, or lower echogenicity compared with that of the adjacent liver parenchyma were defined as hyperechoic, isoechoic, or hypoechoic, respectively. The contrast enhancement pattern (Fig 1) was classified as absent—that is, with no difference in enhancement between the lesion before and the lesion after microbubble contrast agent injection; dotted—that is, with tiny separate spots of enhancement distributed throughout the lesion; rimlike—that is, with a continuous ring of peripheral contrast enhancement; nodular—that is, with discontinuous or continuous peripheral enhancement and a nodular appearance; central—that is, with enhancement of the central portion of the lesion, which was defined as spoke wheel–shaped if a central vessel appeared to branch from the center to the periphery of the lesion; or diffuse—that is, with homogeneous or heterogeneous enhancement of the entire lesion.
View larger version (14K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1. Different contrast enhancement patterns in focal liver lesions. Absent (A), dotted (B), peripheral rimlike (C), peripheral nodular (D), central with spoke wheel-shaped (E), diffuse homogeneous (F), and diffuse heterogeneous (G) enhancement patterns are shown.
|
|
Off-Site Retrospective Analysis
Digital cine clips stored on digital videodisks were reviewed retrospectively on screen by two independent readers (L.G., L.R.) by using Power-DVD software (CyberLink, Fremont, Calif). Both readers were sonologists and had 5 (L.G.) and 3 (L.R.) years of experience in using microbubble contrast agents. They were not involved in the US scanning and were affiliated with one of the two hospitals (San Matteo Hospital). These readers were blinded to the identification, clinical histories, biopsy results, and other imaging findings of the patients. Cine clips were presented in a random order, and any identifying information was masked. Four reading sessions separated by a mean of 3 days were necessary to complete the review of all patient scans. For each lesion, the two readers were asked to give a diagnosis of benign or malignant after reviewing the baseline US scan and then after reviewing the contrast-enhanced US scan, according to the criteria reported in Table 2, which were developed on the basis of enhancement patterns previously described at computed tomography (CT), magnetic resonance imaging (20–23), and contrast-enhanced US (5–10). Both readers used a five-point scale to grade diagnostic confidence: Grade 1 meant definitely benign; grade 2, probably benign; grade 3, indeterminate; grade 4, probably malignant; and grade 5, definitely malignant.
Reference Standards
Table 3 provides information regarding the lesions that is based on reference-standard examination findings. For 317 (191 HCCs, one peripheral cholangiocarcinoma, one epithelioid hemangioendothelioma, 64 metastases, nine hemangiomas, 13 focal nodular hyperplasias, nine hepatocellular adenomas, and 29 macroregenerative nodules) of the 452 lesions, the final diagnosis was obtained at histologic analysis of the biopsy or surgical specimen, which was obtained 24 hours to 20 days after the US examination. With the exception of one fibrolamellar HCC that was detected in a 44-year-old man without cirrhosis, all HCCs and macroregenerative nodules developed secondary to liver cirrhosis. Lesion biopsy revealed 45 poorly differentiated, 102 moderately differentiated, and 43 well-differentiated HCCs; and 20 highly or moderately differentiated and nine dysplastic macroregenerative nodules. Standardized histologic criteria were used to differentiate well-differentiated HCCs from macroregenerative nodules (24). Epithelioid hemangioendothelioma had a low-grade malignant pattern at histologic analysis (25,26).
For the remaining 135 lesions, the final diagnosis was made by using multiphase contrast-enhanced helical CT, which depicted the patterns typical (Table 3) of malignant (41 HCCs and 25 metastases) or benign (47 hemangiomas, 13 focal nodular hyperplasias, and nine regions of focal fatty sparing or change) lesions, followed by serial US performed at least 3–6 months apart with the aid of any relevant clinical information and with increases in lesion diameter and/or number considered criteria for malignancy.
At each hospital involved in the study, the CT images were interpreted by senior radiologists (one of whom was R.P.M.) who had 15–25 years of experience in liver CT and were blinded to the results of all of the other imaging examinations. CT window settings were used according to the local experience and practice of the readers. Established CT criteria (20) were used to characterize the lesions.
Statistical Analyses
Statistical analyses were performed by using a computer software package (Analyse-it, version 1.63; Analyse-it-Software, Leeds, England). The retrospectively determined benign or malignant diagnoses were judged to be true-positive—that is, the lesion was assigned a confidence grade of 4 or 5 and correctly assessed as malignant; false-negative—that is, the lesion was assigned a confidence grade of 1 or 2 and incorrectly assessed as benign or assigned a confidence grade of 3 and judged to be indeterminate; true-negative—that is, the lesion was assigned a confidence grade of 1 or 2 and correctly assessed as benign; or false-positive—that is, the lesion was assigned a confidence grade of 4 or 5 and incorrectly assessed as malignant or assigned a confidence grade of 3 and judged to be indeterminate. Sensitivity was defined as TP/(TP + FN); specificity, as TN/(TN + FP); positive predictive value, as TP/(TP + FP); negative predictive value, as TN/(TN + FN); and overall accuracy, as (TP + TN)/(TP + TN + FP + FN), where TP is the number of true-positive diagnoses; FN, the number of false-negative diagnoses; TN, the number of true-negative diagnoses; and FP, the number of false-positive diagnoses.
To assess the improvement in diagnostic performance after microbubble contrast agent injection for retrospective characterization of each lesion as benign or malignant, the McNemar test (27) was used to evaluate sensitivity and specificity and the 
2 test with Yates correction (27) was used to evaluate positive and negative predictive values and accuracy. The improvement in diagnostic confidence was assessed at receiver operating characteristic curve analysis by using the response from the five-point grade scale and by plotting the sensitivity (ie, true-positive fraction) against 1 – specificity (ie, false-positive fraction). The area under each receiver operating characteristic curve was calculated by using a nonparametric method, as described by Beck and Shultz (28), and was used to indicate the overall diagnostic confidence. Areas under the curves were compared by using a dedicated formula according to the Hanley-McNeil method for analyzing paired data (29). For all tests, P < .05 was considered to indicate a statistically significant difference.
For retrospective analysis, weighted 
statistics were calculated to assess interreader agreement in diagnostic confidence at both review of the baseline US scans and review of the contrast-enhanced US scans. Agreement was graded as poor ( < 0.20), moderate ( 0.20 to <0.40), fair ( 0.40 to <0.60), good ( 0.60 to <0.80), or very good ( 0.80–1.00).
|
RESULTS
|
|---|
On-Site Analysis
HCCs had diffuse homogeneous (70 lesions smaller than 3 cm and 12 lesions larger than 3 cm in diameter) or heterogeneous (12 lesions smaller than 3 cm and 132 lesions larger than 3 cm) contrast enhancement 20–30 seconds after microbubble contrast agent injection. From 45 to 70 seconds after injection to the late phase, HCCs were hypoechoic (n = 137, Fig 2) or isoechoic (n = 89, Fig 3) relative to the adjacent liver, with a peripheral hyperechoic rim in 11 cases. Five additional HCCs that were smaller than 3 cm in diameter had persistently dotted contrast enhancement, with a hypoechoic appearance initially and an isoechoic appearance subsequently during the late phase (Fig 4). The one fibrolamellar HCC had persistent rimlike contrast enhancement and a hypoechoic appearance during the late phase.
View larger version (165K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Transverse contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging in 65-year-old man show most frequent and typical appearances of HCC. (a) Scan obtained during arterial phase—that is, 20 seconds after microbubble contrast agent injection—shows HCC (arrow) with diffuse, slightly heterogeneous contrast enhancement. (b) Scan obtained during portal venous phase—that is, 70 seconds after the injection—shows the HCC (arrow) to be slightly hypoechoic compared with the adjacent liver. (c) Scan obtained during late phase—that is, 100 seconds after the injection—shows the HCC (arrow) to be hypoechoic.
|
|
View larger version (192K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Transverse contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging in 65-year-old man show most frequent and typical appearances of HCC. (a) Scan obtained during arterial phase—that is, 20 seconds after microbubble contrast agent injection—shows HCC (arrow) with diffuse, slightly heterogeneous contrast enhancement. (b) Scan obtained during portal venous phase—that is, 70 seconds after the injection—shows the HCC (arrow) to be slightly hypoechoic compared with the adjacent liver. (c) Scan obtained during late phase—that is, 100 seconds after the injection—shows the HCC (arrow) to be hypoechoic.
|
|
View larger version (174K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2c. Transverse contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging in 65-year-old man show most frequent and typical appearances of HCC. (a) Scan obtained during arterial phase—that is, 20 seconds after microbubble contrast agent injection—shows HCC (arrow) with diffuse, slightly heterogeneous contrast enhancement. (b) Scan obtained during portal venous phase—that is, 70 seconds after the injection—shows the HCC (arrow) to be slightly hypoechoic compared with the adjacent liver. (c) Scan obtained during late phase—that is, 100 seconds after the injection—shows the HCC (arrow) to be hypoechoic.
|
|
View larger version (144K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a. HCC depicted on longitudinal contrast-enhanced US scans obtained in contrast-specific mode with pulse-inversion imaging in 60-year-old man. (a) Scan obtained during arterial phase—that is, 22 seconds after microbubble contrast agent injection—shows HCC (arrow) with diffuse heterogeneous contrast enhancement. (b, c) The tumor (arrow) appears isoechoic relative to the adjacent liver on scans obtained during both portal venous phase—that is, 65 seconds after the injection (b)—and late phase—that is, 100 seconds after the injection (c).
|
|
View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b. HCC depicted on longitudinal contrast-enhanced US scans obtained in contrast-specific mode with pulse-inversion imaging in 60-year-old man. (a) Scan obtained during arterial phase—that is, 22 seconds after microbubble contrast agent injection—shows HCC (arrow) with diffuse heterogeneous contrast enhancement. (b, c) The tumor (arrow) appears isoechoic relative to the adjacent liver on scans obtained during both portal venous phase—that is, 65 seconds after the injection (b)—and late phase—that is, 100 seconds after the injection (c).
|
|
View larger version (148K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3c. HCC depicted on longitudinal contrast-enhanced US scans obtained in contrast-specific mode with pulse-inversion imaging in 60-year-old man. (a) Scan obtained during arterial phase—that is, 22 seconds after microbubble contrast agent injection—shows HCC (arrow) with diffuse heterogeneous contrast enhancement. (b, c) The tumor (arrow) appears isoechoic relative to the adjacent liver on scans obtained during both portal venous phase—that is, 65 seconds after the injection (b)—and late phase—that is, 100 seconds after the injection (c).
|
|
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Longitudinal contrast-enhanced US scans obtained in contrast-specific mode with pulse-inversion imaging in 55-year-old man show less frequent appearances of HCC, which are similar to appearances of macroregenerative nodules. (a) Scan obtained 25 seconds after the injection (arterial phase) shows HCC (arrow) with dotted contrast enhancement, which consists of tiny separate spots, and a hypoechoic appearance. (b, c) Dotted contrast enhancement (arrow) persists during both portal venous phase—that is, 70 seconds after the injection (b)—and late phase—that is, 120 seconds after the injection (c)—with a progressive isoechoic appearance.
|
|
View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Longitudinal contrast-enhanced US scans obtained in contrast-specific mode with pulse-inversion imaging in 55-year-old man show less frequent appearances of HCC, which are similar to appearances of macroregenerative nodules. (a) Scan obtained 25 seconds after the injection (arterial phase) shows HCC (arrow) with dotted contrast enhancement, which consists of tiny separate spots, and a hypoechoic appearance. (b, c) Dotted contrast enhancement (arrow) persists during both portal venous phase—that is, 70 seconds after the injection (b)—and late phase—that is, 120 seconds after the injection (c)—with a progressive isoechoic appearance.
|
|
View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4c. Longitudinal contrast-enhanced US scans obtained in contrast-specific mode with pulse-inversion imaging in 55-year-old man show less frequent appearances of HCC, which are similar to appearances of macroregenerative nodules. (a) Scan obtained 25 seconds after the injection (arterial phase) shows HCC (arrow) with dotted contrast enhancement, which consists of tiny separate spots, and a hypoechoic appearance. (b, c) Dotted contrast enhancement (arrow) persists during both portal venous phase—that is, 70 seconds after the injection (b)—and late phase—that is, 120 seconds after the injection (c)—with a progressive isoechoic appearance.
|
|
The one peripheral cholangiocarcinoma had a persistent heterogeneous and hypoechoic appearance. The one epithelioid hemangioendothelioma had persistent diffuse and homogeneous contrast enhancement and a persistent hyperechoic appearance (Fig 5).
View larger version (48K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Transverse US scans obtained in 66-year-old man show epithelioid hemangioendothelioma with low-grade malignant pattern at histologic analysis. (a) On baseline (left) and color Doppler (right) US scans, the tumor (arrow) appears hypoechoic with a peripheral vessel. (b, c) US scans obtained in contrast-specific mode with pure harmonic detection show diffuse, homogeneous contrast enhancement (arrow) during arterial (b) and portal venous (c) phases. (c) During portal venous phase, the homogeneous enhancement (arrow) persists with a hyperechoic appearance.
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Transverse US scans obtained in 66-year-old man show epithelioid hemangioendothelioma with low-grade malignant pattern at histologic analysis. (a) On baseline (left) and color Doppler (right) US scans, the tumor (arrow) appears hypoechoic with a peripheral vessel. (b, c) US scans obtained in contrast-specific mode with pure harmonic detection show diffuse, homogeneous contrast enhancement (arrow) during arterial (b) and portal venous (c) phases. (c) During portal venous phase, the homogeneous enhancement (arrow) persists with a hyperechoic appearance.
|
|
View larger version (100K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Transverse US scans obtained in 66-year-old man show epithelioid hemangioendothelioma with low-grade malignant pattern at histologic analysis. (a) On baseline (left) and color Doppler (right) US scans, the tumor (arrow) appears hypoechoic with a peripheral vessel. (b, c) US scans obtained in contrast-specific mode with pure harmonic detection show diffuse, homogeneous contrast enhancement (arrow) during arterial (b) and portal venous (c) phases. (c) During portal venous phase, the homogeneous enhancement (arrow) persists with a hyperechoic appearance.
|
|
Metastases were found in patients with the following primary cancers: colorectal (n = 39), gastric (n = 10), or pancreatic (n = 5) adenocarcinoma; squamous cell lung carcinoma (n = 6); breast carcinoma (n = 19); renal cell carcinoma (n = 2); small-bowel carcinoid tumor (n = 2); neuroendocrine pancreatic tumor (n = 2); and cutaneous melanoma (n = 4). During the arterial phase, the metastases had absent (n = 12), dotted (n = 25), diffuse (n = 30), or rimlike (n = 22) contrast enhancement (Fig 6). The diffusely enhancing metastases appeared hyperechoic 20–30 seconds after the microbubble contrast agent injection, and there was evidence of a peripheral hyperechoic rim in the majority (n = 25) of these cases 30–60 seconds after the injection. All metastases appeared hypoechoic and without a peripheral rim during the late phase.
View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6a. Transverse contrast-enhanced US scans obtained in 50-year-old-woman show typical appearances of metastasis. (a) Baseline color Doppler US scan shows slightly heterogeneous metastasis with peripheral vessels. (b, c) Contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging show persistent peripheral, rimlike contrast enhancement 30 seconds (b, arterial phase) and 75 seconds (c, portal venous phase) after microbubble contrast agent injection, with the center of the lesion remaining hypoechoic.
|
|
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6b. Transverse contrast-enhanced US scans obtained in 50-year-old-woman show typical appearances of metastasis. (a) Baseline color Doppler US scan shows slightly heterogeneous metastasis with peripheral vessels. (b, c) Contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging show persistent peripheral, rimlike contrast enhancement 30 seconds (b, arterial phase) and 75 seconds (c, portal venous phase) after microbubble contrast agent injection, with the center of the lesion remaining hypoechoic.
|
|
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6c. Transverse contrast-enhanced US scans obtained in 50-year-old-woman show typical appearances of metastasis. (a) Baseline color Doppler US scan shows slightly heterogeneous metastasis with peripheral vessels. (b, c) Contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging show persistent peripheral, rimlike contrast enhancement 30 seconds (b, arterial phase) and 75 seconds (c, portal venous phase) after microbubble contrast agent injection, with the center of the lesion remaining hypoechoic.
|
|
Most (44 of 56) of the liver hemangiomas had nodular peripheral enhancement during the arterial phase, with progressive centripetal fill-in 50–280 seconds after the microbubble contrast agent injection (Fig 7). Diffuse contrast enhancement with homogeneous fill-in was identified 25–35 seconds after the injection in seven biopsy-proved liver hemangiomas smaller than 3 cm in diameter, which appeared hyperechoic (five hemangiomas) or isoechoic (two hemangiomas) during the late phase. A persistent hypoechoic appearance due to absent contrast enhancement was observed in two liver hemangiomas smaller than 3 cm that were proved to be thrombosed at histologic analysis. Finally, persistent dotted contrast enhancement was seen in three liver hemangiomas smaller than 3 cm.
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. Longitudinal contrast-enhanced US scans obtained in 55-year-old woman show typical appearances of liver hemangioma. (a) Baseline US scan shows hyperechoic and slightly heterogeneous hemangioma (arrow). (b-d) Contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging show nodular peripheral enhancement 35 seconds after microbubble contrast agent injection (b, arterial phase) and progressive centripetal fill-in (arrow in d) during late phase—that is, 110 seconds (c) and 180 seconds (d) after the injection.
|
|
View larger version (122K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. Longitudinal contrast-enhanced US scans obtained in 55-year-old woman show typical appearances of liver hemangioma. (a) Baseline US scan shows hyperechoic and slightly heterogeneous hemangioma (arrow). (b-d) Contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging show nodular peripheral enhancement 35 seconds after microbubble contrast agent injection (b, arterial phase) and progressive centripetal fill-in (arrow in d) during late phase—that is, 110 seconds (c) and 180 seconds (d) after the injection.
|
|
View larger version (137K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7c. Longitudinal contrast-enhanced US scans obtained in 55-year-old woman show typical appearances of liver hemangioma. (a) Baseline US scan shows hyperechoic and slightly heterogeneous hemangioma (arrow). (b-d) Contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging show nodular peripheral enhancement 35 seconds after microbubble contrast agent injection (b, arterial phase) and progressive centripetal fill-in (arrow in d) during late phase—that is, 110 seconds (c) and 180 seconds (d) after the injection.
|
|
View larger version (139K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7d. Longitudinal contrast-enhanced US scans obtained in 55-year-old woman show typical appearances of liver hemangioma. (a) Baseline US scan shows hyperechoic and slightly heterogeneous hemangioma (arrow). (b-d) Contrast-enhanced US scans obtained in contrast-specific mode with contrast-tuned imaging show nodular peripheral enhancement 35 seconds after microbubble contrast agent injection (b, arterial phase) and progressive centripetal fill-in (arrow in d) during late phase—that is, 110 seconds (c) and 180 seconds (d) after the injection.
|
|
Most of the focal nodular hyperplasias—17 of 17 lesions larger than 3 cm in diameter and three of nine smaller than 3 cm—had central spoke wheel–shaped contrast enhancement (Fig 8) 15–20 seconds after the microbubble contrast agent injection. In the following seconds, contrast enhancement appeared diffuse and homogeneous in all except six lesions, which had a central hypoechoic region (Fig 8) that corresponded to the central scar observed at helical CT. The remaining six focal nodular hyperplasias, which were smaller than 3 cm, had persistent and diffuse contrast enhancement and a hyperechoic or isoechoic appearance after the microbubble injection at late phase.
View larger version (83K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8a. Transverse contrast-enhanced US scans obtained in contrast-specific mode with pure harmonic detection in 45-year-old woman show typical spoke wheel-shaped contrast enhancement pattern of focal nodular hyperplasia after microbubble contrast agent injection. (a) Central spoke wheel-shaped contrast enhancement (arrow) is evident 15 seconds after the injection (arterial phase). (b) Contrast enhancement (arrow) is diffuse and homogeneous 25 seconds after the injection (arterial phase). (c) Contrast enhancement is hyperechoic 105 seconds after the injection, with a central hypoechoic region that corresponds to the central scar (arrow) that is evident during the late phase.
|
|
View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8b. Transverse contrast-enhanced US scans obtained in contrast-specific mode with pure harmonic detection in 45-year-old woman show typical spoke wheel-shaped contrast enhancement pattern of focal nodular hyperplasia after microbubble contrast agent injection. (a) Central spoke wheel-shaped contrast enhancement (arrow) is evident 15 seconds after the injection (arterial phase). (b) Contrast enhancement (arrow) is diffuse and homogeneous 25 seconds after the injection (arterial phase). (c) Contrast enhancement is hyperechoic 105 seconds after the injection, with a central hypoechoic region that corresponds to the central scar (arrow) that is evident during the late phase.
|
|
View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8c. Transverse contrast-enhanced US scans obtained in contrast-specific mode with pure harmonic detection in 45-year-old woman show typical spoke wheel-shaped contrast enhancement pattern of focal nodular hyperplasia after microbubble contrast agent injection. (a) Central spoke wheel-shaped contrast enhancement (arrow) is evident 15 seconds after the injection (arterial phase). (b) Contrast enhancement (arrow) is diffuse and homogeneous 25 seconds after the injection (arterial phase). (c) Contrast enhancement is hyperechoic 105 seconds after the injection, with a central hypoechoic region that corresponds to the central scar (arrow) that is evident during the late phase.
|
|
The hepatocellular adenomas (n = 9) had diffuse contrast enhancement 15–20 seconds after the microbubble contrast agent injection and a persistent isoechoic or slightly hyperechoic appearance (Fig 9) during the following 10–20 seconds to the late phase, during which they were isoechoic with homogeneous (seven lesions) or heterogeneous (two lesions) contrast enhancement.
View larger version (71K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Hepatocellular adenoma depicted on transverse US scans obtained in 50-year-old man. (a) Baseline color Doppler US scan shows lesion (arrow) with peripheral vessels. (b-d) Contrast-enhanced US scans obtained in contrast-specific mode with cadence contrast pulse sequencing during arterial phase. The lesion (arrow) has diffuse and homogeneous enhancement 17 seconds after microbubble contrast agent injection (b) and becomes progressively isoechoic relative to the adjacent liver parenchyma 25 seconds (c) and 35 seconds (d) after the injection.
|
|
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Hepatocellular adenoma depicted on transverse US scans obtained in 50-year-old man. (a) Baseline color Doppler US scan shows lesion (arrow) with peripheral vessels. (b-d) Contrast-enhanced US scans obtained in contrast-specific mode with cadence contrast pulse sequencing during arterial phase. The lesion (arrow) has diffuse and homogeneous enhancement 17 seconds after microbubble contrast agent injection (b) and becomes progressively isoechoic relative to the adjacent liver parenchyma 25 seconds (c) and 35 seconds (d) after the injection.
|
|
View larger version (117K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9c. Hepatocellular adenoma depicted on transverse US scans obtained in 50-year-old man. (a) Baseline color Doppler US scan shows lesion (arrow) with peripheral vessels. (b-d) Contrast-enhanced US scans obtained in contrast-specific mode with cadence contrast pulse sequencing during arterial phase. The lesion (arrow) has diffuse and homogeneous enhancement 17 seconds after microbubble contrast agent injection (b) and becomes progressively isoechoic relative to the adjacent liver parenchyma 25 seconds (c) and 35 seconds (d) after the injection.
|
|
View larger version (124K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9d. Hepatocellular adenoma depicted on transverse US scans obtained in 50-year-old man. (a) Baseline color Doppler US scan shows lesion (arrow) with peripheral vessels. (b-d) Contrast-enhanced US scans obtained in contrast-specific mode with cadence contrast pulse sequencing during arterial phase. The lesion (arrow) has diffuse and homogeneous enhancement 17 seconds after microbubble contrast agent injection (b) and becomes progressively isoechoic relative to the adjacent liver parenchyma 25 seconds (c) and 35 seconds (d) after the injection.
|
|
Most (24 of 29) of the macroregenerative nodules had persistent dotted contrast enhancement and a hypoechoic appearance during the arterial phase, followed by an isoechoic appearance during the late phase. The five remaining macroregenerative nodules, which had a dysplastic pattern at histologic analysis, had diffuse contrast enhancement during the first 25–30 seconds of the arterial phase and a hypoechoic appearance during the late phase.
Selected focal fatty sparings (n = 5) and focal fatty changes (n = 4) were detected in patients with known colorectal adenocarcinoma; these lesions had homogeneous contrast enhancement and an isoechoic appearance during the late phase.
Off-Site Retrospective Analysis
After reviewing the contrast-enhanced US scans to classify lesions as benign or malignant, reader 1 (Table 4, Fig 10a) changed the diagnostic confidence score for 292 lesions. For 156 (85 malignant, 71 benign) of these 292 lesions, contrast-enhanced US enabled reader 1 to make a correct diagnosis, whereas for the remaining 136 lesions (95 malignant, 41 benign), reader 1 was more confident in the correct characterization, shifting the diagnostic score from 4 to 5 for malignant lesions and from 2 to 1 for benign lesions. Moreover, at baseline US scan review, reader 1 judged 207 lesions (133 malignant, 74 benign) to be indeterminate; this number was reduced to 29 lesions (23 malignant, six benign) after this reader reviewed the contrast-enhanced US scans.
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10a. Graphs illustrate the increase in diagnostic confidence after review of contrast-enhanced US scans. Receiver operating characteristic curves are plotted to discriminate between benign and malignant focal liver lesions after review of baseline US scans (continuous line) and after review of contrast-enhanced US scans (dotted line) for (a) reader 1 and (b) reader 2. The curves are shown against a diagonal (right) line, which represents a review method with which malignant and benign lesions cannot be differentiated.
|
|
View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10b. Graphs illustrate the increase in diagnostic confidence after review of contrast-enhanced US scans. Receiver operating characteristic curves are plotted to discriminate between benign and malignant focal liver lesions after review of baseline US scans (continuous line) and after review of contrast-enhanced US scans (dotted line) for (a) reader 1 and (b) reader 2. The curves are shown against a diagonal (right) line, which represents a review method with which malignant and benign lesions cannot be differentiated.
|
|
Reader 2 (Table 4, Fig 10b) changed the diagnostic confidence score for 290 lesions after reviewing the contrast-enhanced US scans. For 164 (97 malignant, 67 benign) of these 290 lesions, contrast-enhanced US enabled reader 2 to make a correct diagnosis, whereas for the remaining 126 lesions (86 malignant, 40 benign), this reader was more confident in the correct characterization. Moreover, at baseline US scan review, reader 2 judged 202 lesions (134 malignant, 68 benign) to be indeterminate; this number was reduced to 28 lesions (21 malignant, seven benign) after review of the contrast-enhanced US scans.
At the end of the review session, 59 HCCs with an isoechoic appearance during the late phase were incorrectly judged to be benign by reader 1, 47 HCCs with an isoechoic appearance during the late phase were incorrectly judged to be benign by reader 2, and the epithelioid hemangioendothelioma was incorrectly judged to be benign by both readers. Both thrombosed hemangiomas, however, and three (by reader 1) and five (by reader 2) diffusely enhancing macroregenerative nodules were incorrectly diagnosed as malignant. After review of the contrast-enhanced US scans, interreader agreement increased ( = 0.61 at baseline US vs 0.71 at contrast-enhanced US) and good agreement values (0.60 and <0.80) persisted.
|
DISCUSSION
|
|---|
The use of contrast-specific scanning modes with the air-filled microbubble contrast agent SH U 508A (7–10) has been shown to improve the diagnostic performance of baseline US in the characterization of focal liver lesions. Contrast-enhanced US with SH U 508A is limited owing to two technical factors: First, SH U 508A can be insonated only by using high acoustic power to determine extensive bubble destruction, with emission of a wideband frequency signal. Thus, US scanning has to be performed intermittently because microbubbles are rapidly destroyed; consequently, the signal disappears and there is no chance to use different acoustic views.
Second, compared with low-acoustic-power US scanning, destructive high-acoustic-power scanning produces more tissue harmonics, with a more marked mixture of tissue and microbubble harmonics (11) and the production of a substantial tissue background in which the bubble signal is overimposed. Nondestructive US scanning performed with low acoustic power and perfluorocarbon or sulfur hexafluoride–filled microbubble contrast agents such as SonoVue opens up new possibilities, such as real-time assessment of contrast enhancement with minimal microbubble destruction, the potential to change the acoustic window after microbubble injection, and a more effective suppression of tissue background. The possible disadvantages of nondestructive US scanning, as compared with destructive US scanning, may be the lower signal-to-noise ratio and the narrower dynamic range.
In this study, we found that both diagnostic performance and confidence in benign and malignant liver lesion characterization were improved with use of low-acoustic-power US enhanced with sulfur hexafluoride–filled microbubbles, as compared with the diagnostic performance and lesion characterization achieved with baseline US. Moreover, interreader agreement increased (from = 0.61 to = 0.71) after review of the contrast-enhanced US scans, even though values remained in the range of those for good agreement. These agreement results were based on our analysis of contrast enhancement patterns, which revealed increased reliability and reproducibility of benign and malignant lesion classifications.
The principal difference between benign and malignant lesions that we observed after sulfur hexafluoride contrast agent injection was the difference in their appearances during the late phase at subjective assessment. Malignant lesions appeared to be predominantly hypoechoic, whereas benign lesions appeared to be predominantly hyperechoic or isoechoic relative to the adjacent liver parenchyma. These results confirm previous findings (30) of SH U 508A–enhanced US, which indicated a lower bubble uptake (30) in malignant liver lesions than in benign liver lesions during the late phase. Because SH U 508A has been proved to induce a late liver-specific phase (31), which is probably related to microbubbles pooling in liver sinusoids, it is conceivable that benign lesions such as focal nodular hyperplasia have a similar bubble uptake compared with the surrounding liver because they have a similar sinusoidal architecture.
In the case of the described sulfur hexafluoride contrast agent, which has not been proved to induce a late liver-specific phase, the persistent bubble uptake and the similar microbubble washout between benign lesions and the adjacent liver were probably determined on the basis of the similarity in vascular network—that is, in terms of vessel structure and flow velocity—between these structures. In our series, however, the persistent bubble uptake was also observed in some malignant liver lesions, and some benign lesions did not have bubble uptake, but they had a hypoechoic appearance during the late phase.
With the exception of some HCC lesions that were isoechoic and were incorrectly characterized as benign at retrospective analysis, all HCCs had rapidly increasing diffuse contrast enhancement during the arterial phase and a predominantly hypoechoic appearance during the late phase. It is not clear why some HCCs have the different enhancing behavior. A possible explanation is that low-acoustic-power US enables the circulation of low levels of microbubbles, which produces an isoechoic appearance; however, further studies are necessary to determine whether histologic or vascular differences exist. In five additional HCCs and in most of the macroregenerative nodules, persistent dotted contrast enhancement resembling the hypovascular pattern seen at CT was observed (32).
Only one peripheral cholangiocarcinoma and one epithelioid hemangioendothelioma were observed. The peripheral cholangiocarcinoma was correctly classified as malignant at retrospective analysis, whereas the epithelioid hemangioendothelioma was incorrectly judged to be benign because it was persistently hyperechoic during the late phase.
The metastases had various contrast enhancement patterns during the arterial phase; these patterns ranged from absent to rimlike or diffuse. Evidence of diffuse contrast enhancement of metastases during the arterial phase did not compromise the retrospective diagnosis of these lesions because all of them were overtly hypoechoic during the late phase. Rimlike peripheral enhancement was observed in 47 of 89 metastases; 25 of these 47 lesions were and 22 were not preceded by diffuse contrast enhancement. As also documented in previous studies performed with SH U 508A (5–10), in our series rimlike enhancement was highly specific for malignancy because it was observed only in the metastases and HCCs. Rimlike peripheral enhancement must be distinguished from nodular enhancement, which is typical of liver hemangiomas. Nodular enhancement usually appears during the last seconds of the arterial phase and increases in intensity with a centripetal progression during the portal venous and late phases. Rimlike enhancement, however, appears as a continuous perilesional ring of roughly constant intensity.
In our series, nodular peripheral enhancement with centripetal progression was observed only in the liver hemangiomas and was a reliable pattern of benignity. It should be noticed that in this series nodular enhancement was observed in 44 (79%) of the 56 hemangiomas, whereas it was observed in only 44%–60% of cases in previous series involving the use of SH U 508A and destructive US scanning (5–10). In the present study, real-time US evaluation, facilitated by using nondestructive modes, could have enabled a more complete assessment of this contrast enhancement pattern, even during the first seconds of the arterial phase, when the typical nodular pattern can also be identified in rapidly filling hemangiomas. The hypoechoic appearance of thrombosed hemangiomas during the late phase led to the incorrect characterization of these lesions at retrospective analysis.
The central spoke wheel–shaped pattern, consisting of a central vessel branching from the center to the periphery of the lesion, was identified in the focal nodular hyperplasias. This pattern was previously described at color Doppler US (33) and is highly specific for focal nodular hyperplasia, even though its sensitivity was considered to be low. In this series, the spoke wheel pattern was seen in the majority of the focal nodular hyperplasias during the first seconds after sulfur hexafluoride contrast agent injection. In the hepatocellular adenomas, no vascular pattern was considered specific (34), and contrast-enhanced US enabled the correct characterization of these lesions as benign on the basis of their isoechoic appearance during the late phase.
The results of several previous investigations have shown that both CT and US have poor sensitivity and specificity in the detection and characterization of macroregenerative nodules, which strongly depend on the selection criteria, lesion size, and choice of reference-standard procedures (35,36). In this series, some diffusely enhancing macroregenerative nodules with a dysplastic histologic pattern were incorrectly judged to be malignant because of their hypoechoic appearance during the late phase. However, since the distinction between dysplastic macroregenerative nodules and well-differentiated HCCs is often difficult or impossible, even at histologic assessment (24), this incorrect retrospective characterization could be considered understandable.
Most areas of focal fatty sparing or change were excluded because of their typical US appearance, whereas nine such lesions were incorrectly characterized in patients with known malignancies. Contrast-enhanced US enabled the correct characterization of these lesions as benign.
The principal limitation of this study was the variety of US equipment and contrast-specific modes used. Although this limitation may have been inevitable because this was a multicenter study, the different sensitivities to microbubble harmonics due to the different equipments, transducers, frequencies, and contrast-specific modes should be the subject of further investigation in the future.
A second limitation was the lack of histologic correlation for approximately one-third of the lesions. However, all of the lesions that were not examined at histologic analysis were well characterized at multiphase contrast-enhanced CT on the basis of typical contrast enhancement patterns that are established diagnostic criteria. The lesions that had an atypical appearance or that were not characterized at CT were sampled at biopsy or surgically removed for characterization.
In conclusion, the results of this study show that the use of contrast-specific US modes with a sulfur hexafluoride–filled microbubble contrast agent leads to improved diagnostic performance and improved confidence in characterizing solid focal hepatic lesions as benign or malignant.
|
FOOTNOTES
|
|---|
Abbreviation: HCC = hepatocellular carcinoma
Author contributions: Guarantors of integrity of entire study, E.Q., F.C., M.B.; study concepts and design, E.Q.; literature research, E.Q., M.B.; clinical studies, E.Q.., F.C., S.R., M.B.; data acquisition, E.Q., F.C., S.R., M.B.; data analysis/interpretation, all authors; statistical analysis, E.Q.; manuscript preparation and editing, E.Q.; manuscript definition of intellectual content, E.Q., F.C., M.B.; manuscript revision/review, E.Q., M.B., R.P.M.; manuscript final version approval, E.Q., F.C., M.B., R.P.M.
|
REFERENCES
|
|---|
- Reinhold C, Hammers L, Taylor CR, Quedens-Case CL, Holland CK, Taylor KJ. Characterization of focal hepatic lesions with duplex sonography: findings in 198 patients. AJR Am J Roentgenol 1995; 164:1131-1135.[Abstract/Free Full Text]
- Lee MG, Auh YH, Cho KS, Chung YH, Lee IC, Kang EM. Color Doppler flow imaging of hepatocellular carcinomas: comparison with metastatic tumors and hemangiomas by three step grading color hues. Clin Imaging 1996; 20:199-203.[CrossRef][Medline]
- Hosten N, Puls R, Lemke AJ, et al. Contrast enhanced power Doppler sonography: improved detection of characteristic flow patterns in focal liver lesion. J Clin Ultrasound 1999; 27:107-115.[CrossRef][Medline]
- Kim TK, Han JK, Kim AY, Choi BI. Limitations of characterization of hepatic hemangiomas using an ultrasound contrast agent (Levovist) and power Doppler ultrasound. J Ultrasound Med 1999; 18:737-743.[Abstract]
- Wilson SR, Burns PN, Murdali D, Wilson J, Lai X. Harmonic hepatic ultrasound with microbubble contrast agent: initial experience showing improved characterization of haemangioma, hepatocellular carcinoma, and metastasis. Radiology 2000; 215:153-161.[Abstract/Free Full Text]
- Burns P, Wilson S, Simpson D. Pulse inversion imaging of liver blood flow: improved method for characterization focal masses with microbubble contrast. Invest Radiol 2000; 35:58-71.[CrossRef][Medline]
- Bertolotto M, Dalla Palma L, Quaia E, Locatelli M. Characterization of unifocal liver lesions with pulse inversion harmonic imaging after Levovist injection: preliminary results. Eur Radiol 2000; 10:1369-1376.[CrossRef][Medline]
- Tanaka S, Ioka T, Oshikawa O, Hamada Y, Yoshioka F. Dynamic sonography of hepatic tumors. AJR Am J Roentgenol 2001; 177:799-805.[Abstract/Free Full Text]
- Kim TK, Choi BI, Han JK, Hong HS, Park SH, Moon SG. Hepatic tumors: contrast agent-enhancement patterns with pulse-inversion harmonic US. Radiology 2000; 216:411-417.[Abstract/Free Full Text]
- Dill-Macky M, Burns P, Khalili K, Wilson S. Focal hepatic masses: enhancement patterns with SH U 508 A and pulse-inversion US. Radiology 2002; 222:95-102.[Abstract/Free Full Text]
- Cosgrove DO, Blomley MJ, Eckersley RJ, Harvey C. Innovative contrast specific imaging with ultrasound. Electromedica 2002; 70:147-149.
- Schneider M, Arditi M, Barrau MB, et al. BR1: a new ultrasonographic contrast agent based on sulphur hexafluoride-filled microbubbles. Invest Radiol 1995; 30:451-457.[Medline]
- Morel DR, Schwieger I, Hohn L, et al. Human pharmacokinetics and safety evaluation of SonoVue, a new contrast agent for ultrasound imaging. Invest Radiol 2000; 35:80-85.[CrossRef][Medline]
- Correas JM, Burns PN, Lai X, Qi X. Infusion versus bolus of an ultrasound contrast agent: in vivo dose-response measurements of BR1. Invest Radiol 2000; 35:72-79.[CrossRef][Medline]<
TOP
ABSTRACT
INTRODUCTION
