Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

DOI: 10.1148/radiol.2262011593
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hussain, H. K.
Right arrow Articles by Chenevert, T. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hussain, H. K.
Right arrow Articles by Chenevert, T. L.
(Radiology 2003;226:558-566.)
© RSNA, 2003

Technical Developments

Hepatic Arterial Phase MR Imaging with Automated Bolus-Detection Three-dimensional Fast Gradient-Recalled-Echo Sequence: Comparison with Test-Bolus Method1

Hero K. Hussain, MD, Frank J. Londy, RT, Isaac R. Francis, MD, Hanh V. Nghiem, MD, William J. Weadock, MD, Achamyeleh Gebremariam, MSc and Thomas L. Chenevert, PhD

1 From the Department of Radiology, University of Michigan Health System, 1500 E Medical Center Dr, UH B2B311-MRI, Ann Arbor, MI 48109-0030. Received September 27, 2001; revision requested December 10; final revision received May 8, 2002; accepted May 15. Address correspondence to H.K.H. (e-mail: hhussain@umich.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
Sixty-two patients underwent magnetic resonance (MR) imaging of the liver with the automated contrast material bolus–detection technique. Arterial phase MR images were assessed quantitatively and qualitatively. In 23 patients, a test bolus of contrast material was injected intravenously before dynamic MR imaging. There was good correlation and agreement between delay times estimated with both timing methods. Eighty-three percent of arterial phase images obtained with automated contrast material bolus detection were optimal. There was good correlation and agreement between delay times estimated with both timing methods. Optimal hepatic arterial phase MR images can be obtained routinely with automated detection of a contrast material bolus.

© RSNA, 2003

Index terms: Liver, cirrhosis, 761.794 • Liver, MR, 761.121411, 761.121412 • Liver neoplasms, MR, 761.121411, 761.121412 • Magnetic resonance (MR), contrast enhancement, 761.121411, 761.121412 • Magnetic resonance (MR), vascular studies


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and Methods
Results
Discussion
 REFERENCES
 
Dynamic multiphase magnetic resonance (MR) imaging of the liver has greatly improved the ability to detect and characterize hepatic lesions. The arterial phase is an integral component of dynamic MR imaging because the majority of liver tumors receive their blood supply from the hepatic artery and are most conspicuous in this phase (1,2). This is particularly true for hepatocellular carcinoma, which is found commonly in patients with cirrhosis (3,4). Most hepatocellular carcinomas are hypervascular, and only a minority are hypovascular (5). The duration of the arterial supply predominance in the liver is relatively short and is rapidly predominated by the portal venous supply, which increases the background signal of the liver and obscures lesions that are most conspicuous in the arterial phase. Thus, accurate timing of the arterial phase is crucial to maximize lesion detection and improve characterization.

Two methods—arbitrary fixed delay and test bolus—have been used traditionally to determine the delay time from the start of the injection to arterial phase MR imaging. With the arbitrary fixed-delay method, MR imaging is initiated 15–20 seconds after the start of the injection. This technique does not take into account injection- or patient-related variables. With the test-bolus method, 1–2 mL of contrast medium is injected, and simultaneous rapid and repetitive MR imaging at one level is performed during free breathing to determine the time to peak aortic enhancement. The delay time is calculated on the basis of the time of peak aortic enhancement, injection volume, and rate. While considered the most accurate, this method requires additional imaging and calculations. Furthermore, delay errors up to 6 seconds have been documented when test-bolus and dynamic MR imaging are performed in different phases of respiration (6).

The automated contrast material bolus– detection technique takes into account injection- and patient-related variables. This method has been used successfully for the timing of MR angiographic acquisitions but has not been applied widely for the timing of dynamic liver MR imaging.

The purpose of our study was to assess the automated bolus-detection technique for providing arterial phase MR images of the liver.


   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
Patients
Between October and December 2000, 65 consecutive patients underwent dynamic hepatic MR imaging. We received institutional review board approval for our study, and they waived the need for informed consent. Three patients were excluded from the study because of their inability to suspend respiration (n = 2) and power injector failure (n = 1). Therefore, the study group comprised 62 patients (41 men and 21 women; mean age, 57 years; age range, 28–85 years). Referring diagnoses were suspected hepatocellular carcinoma (n = 31), indeterminate hepatic lesion at ultrasonography (US) or computed tomography (CT) (n = 12), suspected metastases (n = 12), posttransplantation lymphoproliferative disease (n = 5), and suspected cholangiocarcinoma (n = 2). Of the 62 patients, 27 had known cirrhosis and portal hypertension, four had cirrhosis without radiologic evidence of portal hypertension, and 31 did not have cirrhosis. The presence or absence of cirrhosis was determined at liver biopsy in all cases.

MR Imaging
MR imaging was performed with a 1.5-T MR system (Signa; GE Medical Systems, Milwaukee, Wis) with high-performance gradients (maximum gradient strength, 25 mT/m; rise time, 600 seconds) and a torso phased-array coil. Before MR imaging, a 20–24-gauge needle was placed in the antecubital fossa and attached to an MR-compatible power injector (Spectris; Medrad, Pittsburgh, Pa). All but two patients received 20 mL of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ), one patient received 30 mL, and one received 40 mL. In all patients, a 15-mL saline flush followed the contrast material bolus. Both contrast material and saline were administered at a rate of 2 mL/sec.

All patients underwent a routine liver MR examination that included the following sequences: breath-hold coronal T2-weighted single-shot fast spin echo, transverse T1-weighted in-phase and out-of-phase gradient-recalled echo, transverse respiratory-triggered T2-weighted fast spin echo, and pre- and postcontrast three-phase dynamic automated bolus-detection three-dimensional fast gradient-recalled echo.

Automated bolus-detection three-dimensional spoiled fast gradient-recalled echo.—Symmetric k-space sampling in the section-select partial–k-space sampling in the readout (ie, asymmetric echo) and phase-encoding (ie, partial Fourier) directions was used.

Dynamic liver MR imaging was performed with the three-dimensional fast gradient-recalled-echo sequence with the following parameters: 4–6/<2 (repetition time msec/echo time msec), flip angle of 12°, bandwidth of 31.25 kHz, and spectral fat saturation (inversion time of 20–30 msec). Automated bolus detection (SmartPrep; GE Medical Systems) is an optional choice for the three-dimensional fast gradient-recalled-echo sequence that was turned on for the dynamic study in all cases. The automated bolus-detection technique has been described fully in previous publications (7,8). The contrast material detection volume, or "tracker," of the automated bolus-detection method was placed on the aorta at the level of the celiac axis on the most suitable image obtained with the coronal single-shot fast spin-echo sequence (Fig 1). In all patients, the fail-safe mechanism, also called the maximum monitor period, was set to 35 seconds with a delay time of 8 seconds between contrast material detection and initiation of data acquisition. Centric k-space encoding is used with the automated contrast material bolus–detection technique to initiate acquisition of central k-space data at the desired time in peak arterial enhancement.



View larger version (126K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Coronal T2-weighted single-shot fast spin-echo MR image (21,998/95) through the aorta shows position of tracker volume for automated bolus detection over abdominal aorta at level of celiac axis. Note tracker volume (arrow) is slightly wider than aortic lumen.

 
A precontrast three-dimensional fast gradient-recalled-echo sequence was performed initially with the automated bolus-detection option turned off to ensure adequate coverage and provide a baseline study. This was followed with three dynamic acquisitions, in the arterial-dominant, portal venous, and equilibrium phases; each acquisition necessitated a 16–29-second breath hold. Patients were instructed to relax for approximately 20 seconds between these phases. The delay to arterial phase MR imaging that was determined with the automated method (contrast material detection time + 8 seconds) was recorded.

The imaging plane was transverse in 59 patients and coronal in three. A rectangular field of view was used for transverse MR imaging to reduce imaging time, and a square field of view was used for coronal MR imaging to reduce aliasing artifacts. Partial (0.5) excitation was used in the majority of patients to reduce imaging time, but full excitation was preferred when possible, usually in patients with small livers and those able to suspend respiration for 26–29 seconds. A section thickness of 4–5 mm and a matrix of 320 x 160 (frequency x phase) were used routinely except when patients were unable to hold their breath. In these cases, the matrix was reduced to 256–288 x 128. Normally, k space was filled with 320 points along kx, 160 points along ky, and 40 points along kz. Zero interpolation was performed in the section-select direction to approximately double the number of sections and in the read-out and phase directions to a 512 x 512 matrix. Slab thickness was 160–200 mm divided into 80 partitions (after interpolation), which resulted in effective partition spacing of 2.0–2.5 mm. A field of view of 320–400 mm yielded a voxel size (true resolution) of approximately 1 x 2 x 4 mm (maximum, 1.25 x 2.25 x 5.00 mm) and an image pixel size of 0.6 x 0.6 x 2.0 (maximum, 0.8 x 0.8 x 2.5).

Test bolus.—Twenty-three consecutive patients underwent a test-bolus examination before automated bolus-detection dynamic MR imaging (Table 1). The test bolus was injected during free breathing. Gadopentetate dimeglumine (2 mL) followed by a 15-mL saline flush were injected at a rate of 2 mL/sec. A multiphase single-level transverse T1-weighted MR image of the aorta at the level of the celiac axis was obtained every 2 seconds from the start of injection for 60 seconds with a two-dimensional spoiled gradient-recalled-echo sequence (20/1.6 [epetition time msec/echo time msec], flip angle of 30°, section thickness of 10 mm). Superior and inferior saturation bands were used to reduce time-of-flight enhancement of intravascular signal and to maximize depiction of the arrival of contrast material.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Predicted and Calculated Delay Times to Arterial Phase MR Imaging Determined with Test-Bolus and Automated Methods

 
One investigator (H.K.H.) placed an approximately 2–3-cm2 region of interest to fit within the aortic lumen on the MR image that showed maximal aortic enhancement. Curves for signal intensity (SI) to section number were generated by means of vendor-provided software (Functool on Advantage Workstation; GE Medical Systems). The time to peak aortic enhancement was estimated with the knowledge that each section was acquired in 2 seconds. The optimal imaging delay (OID) to arterial phase MR imaging (ie, the delay time from the beginning of injection to the beginning of MR imaging) was calculated for a centric k-space–encoded sequence by means of the following formula (9): OID = Dpae + (tinj/2), where Dpae is the delay to peak aortic enhancement and tinj is the time for contrast material injection. The time for the contrast material injection was calculated by dividing the volume of contrast material by the rate of injection.

Data Analysis
Quantitative measurement on arterial phase MR images.—One investigator (H.K.H.) placed a region of interest on all the automated bolus-detection three-dimensional fast gradient-recalled-echo MR images. Regions of interest that were 0.5–8.0 cm2 were placed over the following areas in all four phases (ie, precontrast, arterial-dominant, portal venous, and delayed): liver, aorta, suprarenal inferior vena cava, hepatic artery, portal vein, hepatic vein, spleen, renal cortex, and noise (measured in air outside the patient’s body in the phase-encoding direction, either anterior or lateral to the body on transverse and coronal images, respectively, avoiding phase artifacts).

Four large regions of interest (typically 4–8 cm2) were placed over different parts of the right and left lobes of the liver with care to avoid vascular structures. The mean of these measurements was used to represent liver parenchymal enhancement during any of the four phases. The mean SI values and SDs were used to calculate parameters similar to those used in other studies (10,11) to determine the adequacy of arterial phase MR images and the tissue and vessel enhancement obtained in all phases.

The following parameters were calculated from these measures. (a) Signal-to-noise ratio (SNR) of target tissue for each phase: SNRtissue = SItissue/SDnoise. (b) Percentage liver enhancement (%LEliver) in the arterial-dominant phase compared with peak parenchymal enhancement: %LEliver = [(SIart - SIpre)/(SIpeak - SIpre)] x 100, where SIpre is in the precontrast phase, SIart is in the arterial phase, and SIpeak is in the phase during which there was maximal hepatic parenchymal enhancement, either portal venous or delayed.

Others (12) have proposed that the arterial phase MR images are considered to be optimally timed if the percentage liver parenchymal enhancement in the arterial phase is equal to or less than 30% of peak parenchymal enhancement, which is seen in the portal venous or delayed phases. While this criterion is appropriate in patients with normal livers, the relative arterial phase enhancement may be higher in patients with chronic liver disease, cirrhosis, and portal hypertension because of diminished portal venous supply and increased arterial reserve (13,14).

The following parameter was also calculated: (c) venous-to-arterial enhancement ratio (V/A) of the hepatic vein (hv) relative to the hepatic artery (ha): V/A = [SIart(hv) - SIpre(hv)]/[SIart(ha) - SIpre(ha)], where art is in the arterial-dominant phase, and pre is in the precontrast phase. A V/A of 0 represents a well-timed arterial phase with no hepatic venous enhancement.

Qualitative measurement on arterial phase MR images.—Consensus qualitative analysis to determine the adequacy of the arterial phase images was performed by two experienced radiologists (I.R.F., H.V.N.) with use of a point scale of 1–3 (1 = early, 2 = appropriate, and 3 = late). Early arterial phase MR images (Fig 2a) had maximal aortic and hepatic arterial enhancement, no portal or hepatic venous enhancement, and minimal or no splenic, pancreatic, or renal cortical enhancement. Appropriate arterial phase MR images (Fig 2b, 2c) had maximal aortic and hepatic arterial enhancement; mild to moderate portal venous enhancement; no hepatic venous enhancement; and heterogeneous splenic, uniform pancreatic, and renal cortical enhancement. Late arterial phase MR images (Fig 2d, 2e) had maximal aortic and hepatic arterial enhancement, moderate to high portal and hepatic venous enhancement, and uniform splenic, pancreatic, and nephrographic renal enhancement.



View larger version (139K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Transverse subvolume maximum intensity projection of automated bolus-detection three-dimensional fast gradient-recalled-echo sequence (6/2, flip angle of 12°) MR images in three patients: (a) patient 1, early; (b, c) patient 2, appropriate; and (d, e) patient 3, late arterial phases. (a) In early arterial phase, images show (b) maximal aortic and hepatic arterial enhancement (solid arrow) and no portal venous enhancement (arrowhead). Note lack of splenic enhancement (open arrow). (b, c) In appropriate arterial phase, MR images show (b) maximal aortic and hepatic arterial enhancement (solid straight arrow) and mild portal venous enhancement (large arrowhead), heterogeneous splenic enhancement (open arrow), uniform pancreatic enhancement (small arrowhead), and renal cortical enhancement (curved arrow). (c) In appropriate arterial phase, MR image shows no hepatic venous enhancement (arrow). (d, e) In late arterial phase, MR images show (d) maximal aortic and hepatic arterial enhancement (arrow), moderate portal venous enhancement (arrowhead), and (e) moderate hepatic venous enhancement (arrow).

 


View larger version (149K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Transverse subvolume maximum intensity projection of automated bolus-detection three-dimensional fast gradient-recalled-echo sequence (6/2, flip angle of 12°) MR images in three patients: (a) patient 1, early; (b, c) patient 2, appropriate; and (d, e) patient 3, late arterial phases. (a) In early arterial phase, images show (b) maximal aortic and hepatic arterial enhancement (solid arrow) and no portal venous enhancement (arrowhead). Note lack of splenic enhancement (open arrow). (b, c) In appropriate arterial phase, MR images show (b) maximal aortic and hepatic arterial enhancement (solid straight arrow) and mild portal venous enhancement (large arrowhead), heterogeneous splenic enhancement (open arrow), uniform pancreatic enhancement (small arrowhead), and renal cortical enhancement (curved arrow). (c) In appropriate arterial phase, MR image shows no hepatic venous enhancement (arrow). (d, e) In late arterial phase, MR images show (d) maximal aortic and hepatic arterial enhancement (arrow), moderate portal venous enhancement (arrowhead), and (e) moderate hepatic venous enhancement (arrow).

 


View larger version (141K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c. Transverse subvolume maximum intensity projection of automated bolus-detection three-dimensional fast gradient-recalled-echo sequence (6/2, flip angle of 12°) MR images in three patients: (a) patient 1, early; (b, c) patient 2, appropriate; and (d, e) patient 3, late arterial phases. (a) In early arterial phase, images show (b) maximal aortic and hepatic arterial enhancement (solid arrow) and no portal venous enhancement (arrowhead). Note lack of splenic enhancement (open arrow). (b, c) In appropriate arterial phase, MR images show (b) maximal aortic and hepatic arterial enhancement (solid straight arrow) and mild portal venous enhancement (large arrowhead), heterogeneous splenic enhancement (open arrow), uniform pancreatic enhancement (small arrowhead), and renal cortical enhancement (curved arrow). (c) In appropriate arterial phase, MR image shows no hepatic venous enhancement (arrow). (d, e) In late arterial phase, MR images show (d) maximal aortic and hepatic arterial enhancement (arrow), moderate portal venous enhancement (arrowhead), and (e) moderate hepatic venous enhancement (arrow).

 


View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d. Transverse subvolume maximum intensity projection of automated bolus-detection three-dimensional fast gradient-recalled-echo sequence (6/2, flip angle of 12°) MR images in three patients: (a) patient 1, early; (b, c) patient 2, appropriate; and (d, e) patient 3, late arterial phases. (a) In early arterial phase, images show (b) maximal aortic and hepatic arterial enhancement (solid arrow) and no portal venous enhancement (arrowhead). Note lack of splenic enhancement (open arrow). (b, c) In appropriate arterial phase, MR images show (b) maximal aortic and hepatic arterial enhancement (solid straight arrow) and mild portal venous enhancement (large arrowhead), heterogeneous splenic enhancement (open arrow), uniform pancreatic enhancement (small arrowhead), and renal cortical enhancement (curved arrow). (c) In appropriate arterial phase, MR image shows no hepatic venous enhancement (arrow). (d, e) In late arterial phase, MR images show (d) maximal aortic and hepatic arterial enhancement (arrow), moderate portal venous enhancement (arrowhead), and (e) moderate hepatic venous enhancement (arrow).

 


View larger version (137K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2e. Transverse subvolume maximum intensity projection of automated bolus-detection three-dimensional fast gradient-recalled-echo sequence (6/2, flip angle of 12°) MR images in three patients: (a) patient 1, early; (b, c) patient 2, appropriate; and (d, e) patient 3, late arterial phases. (a) In early arterial phase, images show (b) maximal aortic and hepatic arterial enhancement (solid arrow) and no portal venous enhancement (arrowhead). Note lack of splenic enhancement (open arrow). (b, c) In appropriate arterial phase, MR images show (b) maximal aortic and hepatic arterial enhancement (solid straight arrow) and mild portal venous enhancement (large arrowhead), heterogeneous splenic enhancement (open arrow), uniform pancreatic enhancement (small arrowhead), and renal cortical enhancement (curved arrow). (c) In appropriate arterial phase, MR image shows no hepatic venous enhancement (arrow). (d, e) In late arterial phase, MR images show (d) maximal aortic and hepatic arterial enhancement (arrow), moderate portal venous enhancement (arrowhead), and (e) moderate hepatic venous enhancement (arrow).

 
Cirrhosis versus no cirrhosis.—Parenchymal enhancement in all phases, percentage parenchymal enhancement in the arterial phase compared with peak enhancement, and time to peak enhancement were compared between the group of patients with cirrhosis and that without it.

Statistical Analysis
In the 23 patients who underwent test-bolus prior to dynamic liver MR imaging with automated contrast material bolus detection, the optimal imaging delay to arterial phase MR imaging predicted with the test-bolus method and that determined with the automated bolus-detection method were compared by means of the Pearson correlation coefficient. Values closer to 1 indicate strong positive linear association, while those closer to -1 indicate strong negative linear association. Agreement between the two methods was also determined with the Bland and Altman method (15). With this method, agreement is assessed graphically by means of plotting the between-methods difference (bias) against the mean measurement of the two methods. If all or most points are within 2 SDs of the average bias, the agreement between the two methods is considered good.

Repeated measurement of analysis of variance was performed to examine the overall phase effect on hepatic parenchymal enhancement, adjusted for disease group (with cirrhosis vs without cirrhosis). A P value of less than .05 was considered to indicate a statistically significant difference. If significant overall effect was detected, further examination by means of multiple comparisons with Tukey-Kramer adjustment was performed to determine the differences that contribute to the overall significance (SAS Institute. SAS/STAT user’s guide, version 8. Chapter 41. Cary, NC: SAS Institute, 2000).


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
Delay Time
There was good correlation (r = 0.80, P < .001) between the imaging delay time predicted with the test-bolus method and that observed with the automatic-detection method (Fig 3a). Good agreement between the two methods was demonstrated with a negligible average bias of the automated method relative to the test-bolus method, as explained in Figure 3b. The imaging delay, quantitative and qualitative analyses data, and disease condition of the 23 patients who underwent both tests are detailed in Table 1.



View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. (a) Scatterplot of optimal delay to arterial phase predicted with test-bolus method versus that observed with automated bolus-detection method. Good correlation (r = 0.80, P < .001) between methods is demonstrated. (b) Scatterplot of difference in delay times versus average of delay times calculated with both methods in each patient. The difference, which is an estimate of average bias of automated delay relative to test-bolus delay, was 0.22 seconds (SD = 2.6 seconds). All points are within 2 SDs of average bias. This indicates a negligible average bias of automated bolus-detection method relative to test-bolus method and good agreement between the two methods (95% CI: -5.0, 5.4). With both methods, measurements of delay times to arterial phase imaging differed by less than 5 seconds with any discrepancy being equally likely in either direction.

 


View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. (a) Scatterplot of optimal delay to arterial phase predicted with test-bolus method versus that observed with automated bolus-detection method. Good correlation (r = 0.80, P < .001) between methods is demonstrated. (b) Scatterplot of difference in delay times versus average of delay times calculated with both methods in each patient. The difference, which is an estimate of average bias of automated delay relative to test-bolus delay, was 0.22 seconds (SD = 2.6 seconds). All points are within 2 SDs of average bias. This indicates a negligible average bias of automated bolus-detection method relative to test-bolus method and good agreement between the two methods (95% CI: -5.0, 5.4). With both methods, measurements of delay times to arterial phase imaging differed by less than 5 seconds with any discrepancy being equally likely in either direction.

 
Quantitative Measurement of Arterial Phase Images
Signal-to-noise ratio.—The mean signal-to-noise ratio of abdominal organs and vessels in all phases is shown in Table 2. There was significant overall phase effect on hepatic parenchymal enhancement (P < .001). Results of further analyses showed that the mean enhancement values were 3.6 units higher in the arterial-dominant phase compared with those in the precontrast phase (P < .001) and 17.1 units higher in the portal venous phase compared with those in the arterial phase (P < .001), but no significant decline was seen between the portal venous and delayed phases (P = .28) (Table 3).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Signal-to-Noise Ratios of Abdominal Organs and Vessels in All Phases (n = 62)

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Pairwise Comparisons of SI between Phases

 
Percentage liver enhancement in arterial phase.—Mean hepatic enhancement in the arterial phase was 17% ± 13 of peak parenchymal enhancement. Of the 62 patients, 52 (84%) (26 with cirrhosis and 26 without) had relative enhancement values of 30% or less, and 10 (16%) (five with cirrhosis and five without) had values of more than 30% (range, 31%–50%) (Table 4).


View this table:
[in this window]
[in a new window]

  		TABLE 4. SI of Liver in All Four Phases

 
Venous-to-arterial enhancement ratio.—Of the 62 patients, 53 (85%) (26 with cirrhosis and 27 without) had a venous-to-arterial enhancement ratio of less than 0.1, which effectively represents no venous enhancement. Nine (15%) patients (five with cirrhosis and four without) had a ratio in the range of 0.1–1.2, which represents minimal venous enhancement. One patient with tricuspid regurgitation had a ratio of 2.2.

Qualitative Measurement of Arterial Phase MR Images
In the 62 patients, 50 (81%) arterial phase images (25 had cirrhosis and 25 did not) were rated as appropriate by means of consensus qualitative analysis and, 12 (19%) were rated as inappropriate (nine late and three early). Of the nine late images, four had cirrhosis and five did not; of the three early images, two had cirrhosis and one did not.

Cirrhosis versus No Cirrhosis
Thirty-one patients had liver cirrhosis, and 31 did not. No significant difference in parenchymal enhancement was found in SI in any of the phases between groups (P > .35). Mean signal-to-noise ratio for each phase for each group is shown in Table 5. Mean percentage liver enhancement in the arterial phase was 18% ± 13 (with cirrhosis, 17% ± 14; without cirrhosis, 20% ± 12). Peak hepatic parenchymal enhancement was in the portal venous phase in 10 (32%) patients with cirrhosis and 25 (81%) without and in the delayed phase in 16 (52%) patients with cirrhosis and six (19%) without. Five (16%) patients with cirrhosis had similar parenchymal enhancement values in the portal venous and delayed phases.


View this table:
[in this window]
[in a new window]

  		TABLE 5. Mean Signal-to-Noise Ratio of Liver Parenchyma in Each Phase

 

   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
REFERENCES
 
With the automated bolus-detection method for dynamic liver MR imaging, 83% of our study group had optimal arterial phase images according to previously established quantitative criteria (10–12) and qualitative assessment by two experienced radiologists. This method has been used successfully for the timing of MR angiographic studies but not widely applied for abdominal imaging (7). The principle behind automatic detection is that contrast material in a vessel is detected when the SI in that vessel increases 20% above the precontrast level (8). Once contrast material is detected, centric k-space data acquisition is initiated after a delay determined by the operator to ensure maximal vascular enhancement and avoid artifacts (16).

The term optimal arterial phase requires definition; authors have used various quantitative and qualitative criteria to determine it (10–12,17–19). We used two quantitative criteria from prior studies (10–12): (a) parenchymal enhancement in the arterial phase of 30% or less compared with the peak hepatic parenchymal enhancement value and (b) the hepatic venous to hepatic arterial enhancement ratio. The 30% limit for optimal parenchymal enhancement in the arterial phase was established with CT in a group of patients with no underlying liver disease (12). An optimal venous-to-arterial enhancement ratio of 0 indicates the absence of hepatic venous enhancement and is considered to be an indicator of good arterial phase images with no venous contamination. In our study, 84% of the patients fulfilled the first criterion, and 85% fulfilled the second criterion. One patient had tricuspid regurgitation, which resulted in considerable hepatic venous enhancement and a venous-to-arterial enhancement ratio of 2.2.

The timing of the arterial phase is affected by many variables; some are related to the injection, such as rate and volume, and others are related to the patient, such as cardiac output and weight (20–24). Unlike the widely used fixed-delay method, the automated method takes into account patient- and injection-related variables. It is easy to use and does not necessitate the further MR imaging and contrast material injection required with the test-bolus method. Furthermore, because central k space is acquired at the beginning of the sequence, patients who are unable to suspend respiration can be instructed to hold their breath during only the early part of MR imaging, when the critical data in central k space are acquired, without substantially affecting the image quality.

Results of studies to optimize arterial phase images have been variable (10,25,26). To determine the delay to arterial phase MR imaging, Earls et al (10) compared the adequacy of arterial phase MR imaging with the arbitrary fixed-delay and test-bolus methods in two groups of patients. They found that the arterial phase images were adequate in 17 (61%) of 28 patients with the fixed-delay method compared with 26 (93%) of 28 patients with the test-bolus method. Materne et al (26) compared arterial phase images obtained after a fixed delay with those obtained with automated bolus detection in the same group of patients, who were known to have hypervascular tumors and normal cardiac function. They found no significant (P < .05) improvement in tumor-to-liver contrast and enhancement with either method. While this may be true in patients with no underlying liver disease, it may not apply to patients with cirrhosis, who, in addition to abnormal liver function, have hemodynamic changes that affect both the hepatosplanchnic and central circulations. These changes result in increased cardiac output and peripheral vasodilation (27). Furthermore, it is in this group of patients that arterial phase MR imaging is particularly important because the number and size of hepatocellular tumors will determine management and prognosis.

Thirty-one (50%) of the 62 patients in the present study had cirrhosis with or without portal hypertension. Of the 10 patients who had more than 30% enhancement in the arterial phase compared with peak enhancement, five had cirrhosis and portal hypertension. It is well known that disturbances in the blood supply to the liver occur in these patients. Findings in studies of cirrhotic livers have shown that there is arterialization of the liver as a result of enlargement of the arteries and arterial bed. Hepatic venous capillaries become obliterated secondary to fibrosis and pressure from regenerating nodules; the portal veins become narrow and angular. The portal vein in these patients can be converted into an outflow of the liver, and the liver can become totally dependent on the blood supply from the hepatic artery (13,14). Thus, increased arterial supply, rather than the technique, may have been the cause of increased parenchymal enhancement in these patients. The 30% limit on optimal parenchymal enhancement in the arterial phase established at CT in a group of healthy individuals (12) may not apply to patients with cirrhosis, who constituted 50% of our study population. Nevertheless, our results show that this technique is equally effective in patients with and those without cirrhosis, despite the hemodynamic disturbances associated with cirrhosis. There were almost equal numbers of patients with and those without cirrhosis among those with results that did not fulfill our quantitative and qualitative criteria for optimal arterial phase images.

The pattern of hepatic parenchymal enhancement is also influenced by cirrhosis and portal hypertension. In a study by Soyer et al (28) of the patterns of hepatic parenchymal enhancement in 20 patients with cirrhosis and portal hypertension with and without splenomegaly and 20 control subjects with no liver disease, peak hepatic parenchymal enhancement was significantly (P < .05) reduced in patients with portal hypertension and splenomegaly. In addition, time to peak parenchymal enhancement was delayed significantly (P < .01) when portal hypertension was present. In our study group, the mean peak hepatic parenchymal enhancement value in patients with cirrhosis was 48 ± 16 compared with 52 ± 25 in the group without cirrhosis, with no statistically significant difference. These results are similar to those reported by Lee et al (11) of no significant difference between peak parenchymal enhancement values in patients with and those without cirrhosis. Similar to results in the study by Soyer et al (28), we found that the time to peak parenchymal enhancement was delayed in patients with cirrhosis (48% in the portal venous phase, 52% in the delayed phase) compared with that in patients without cirrhosis (81% in the portal venous phase, 19% in the delayed phase). Further studies are necessary to determine the duration of the arterial phase and the effect of portal hypertension in patients with cirrhosis and those with chronic liver disease.

There are pitfalls with the automated bolus-detection technique, including the risk that the patient will move between images and displace the tracker from the desired vessel; the possibility that contrast material will not be detected; and the operator-dependent problems, such as placement of the tracker over the inferior vena cava instead of over the aorta. It is always safer to choose a tracker that is slightly wider than the vessel diameter to reduce the effect of patient motion. In our study, there were no cases of failure to detect contrast material. Another factor that may influence the timing of the arterial phase is the choice of delay time between contrast material detection and initiation of MR imaging. We chose 8 seconds on the basis of trial and error because a 5-second delay is used for arterial imaging in MR angiography (7). We believe that 8 seconds will ensure adequate hepatic arterial opacification and optimal parenchymal enhancement. While this may be slightly delayed, being late is safer than being early because MR imaging in the late arterial phase has been shown to improve hypervascular tumor detection (18,19).

Many authors agree that the optimal phase for hypervascular tumor detection is the late arterial phase, also called the portal venous inflow phase or sinusoidal phase (17–19). Images obtained in this phase have intense hepatic arterial, substantial portal venous, slight parenchymal, and no hepatic venous enhancement, and they are obtained approximately 30 seconds after initiation of the contrast material injection (17). This is somewhat comparable to the enhancement phase obtained with our technique at an average delay of 25 seconds after detection of contrast material in the vessel. Results in a recent study by Murakami et al (17) showed improved detection of hepatocellular carcinoma with CT on early and late arterial phase MR imaging.

Van Beers et al (29) assessed timing optimization for arterial phase MR imaging and found that the highest tumor-to-liver contrast-to-noise ratio is achieved when timing is based on the enhancement profile in the tumor rather than that in the aorta; however, the contrast-to-noise ratio at peak tumor enhancement did not differ significantly from that obtained after peak aortic enhancement (P = .471). In our study, we used the peak aortic enhancement as our guide for the timing of the arterial phase. Automated timing on the basis of tumor enhancement may be possible if the tumor is large, but it can be difficult with small tumors. Furthermore, automated bolus detection may not be sensitive to the minor changes in SI within the tumor. Tumors can be used as guidelines with the test-bolus method or with the MR fluoroscopic triggering techniques that are becoming available commercially.

There are limitations to our study. First, it would have been preferable to perform this comparison with two sets of dynamic MR images in the same patient, one set timed with the test-bolus method and the second set timed with the automated bolus-detection method. This would have necessitated exposure of the patient to two injections and two MR imaging examinations. We believed this was unnecessary because the aim of the study was to evaluate the automated bolus-detection method. Second, we performed the test-bolus method with a limited number of our patients. Third, our choice of delay time between contrast material detection and MR imaging with the automated method was based on our experience, and it had a major influence on our results. Fourth, there are no clear criteria with which to determine the adequacy of arterial phase MR images, and our results might have been influenced by the criteria we chose. Finally, we did not evaluate the sequence for image quality and artifacts.

In conclusion, we find automated detection of a contrast material bolus to be an easy and reliable method that can be used routinely to obtain adequate three-dimensional arterial phase images of cirrhotic and noncirrhotic livers. The technique is patient specific and eliminates the need to guess or to administer a test bolus.


   ACKNOWLEDGMENTS
 
The authors thank Kathie Helm and Robert Combs for their excellent support.


   FOOTNOTES
 
Abbreviation: SI = signal intensity

Author contributions: Guarantor of integrity of entire study, H.K.H.; study concepts and design, H.K.H., T.L.C.; literature research, H.K.H.; clinical studies, H.K.H., I.R.F., H.V.N., W.J.W.; data acquisition, H.K.H., F.J.L., I.R.F., H.V.N., W.J.W.; data analysis/interpretation, H.K.H., F.J.L., T.L.C.; statistical analysis, A.G., F.J.L.; manuscript preparation, H.K.H.; manuscript definition of intellectual content, H.K.H., T.L.C.; manuscript editing, manuscript revision/review, and final version approval, all authors.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 

  1. Bonaldi VM, Bret PM, Reinhold C, Atri M. Helical CT of the liver: value of an early hepatic arterial phase. Radiology 1995; 197:357-363.[Abstract/Free Full Text]
  2. Hollett MD, Jeffrey RB, Jr, Nino-Murcia M, Jorgensen MJ, Harris DP. Dual-phase helical CT of the liver: value of arterial phase scans in the detection of small (< or = 1.5 cm) malignant hepatic neoplasms. AJR Am J Roentgenol 1995; 164:879-884.[Abstract/Free Full Text]
  3. Baron RL, Oliver JH, Dodd GD, III, Nalesnick M, Holbert BL, Carr B. Hepatocellular carcinoma: evaluation with biphasic, contrast-enhanced helical CT. Radiology 1996; 199:505-511.[Abstract/Free Full Text]
  4. Kim T, Murakami T, Takahashi S, et al. Optimal phases of dynamic CT for detecting hepatocellular carcinoma: evaluation of unenhanced and triple-phase images. Abdom Imaging 1999; 24:473-480.[CrossRef][Medline]
  5. Kelekis NL, Semelka RC, Worawattanakul S, et al. Hepatocellular carcinoma in North America: a multiinstitutional study of appearance on T1-weighted, T2-weighted, and serial gadolinium-enhanced gradient-echo images. AJR Am J Roentgenol 1998; 170:1005-1013.[Abstract/Free Full Text]
  6. Krinsky GA, Kaminer E, Lee VS, Rofsky NM, Weinreb JC. The effects of apnea on timing examinations for optimization of gadolinium-enhanced MRA of the thoracic aorta and arch vessels. J Comput Assist Tomogr 1998; 22:677-681.[CrossRef][Medline]
  7. Prince MR, Chenevert TL, Foo TKF, Londy FJ, Ward JS, Maki JH. Contrast-enhanced abdominal MR angiography: optimization of imaging delay time by automating the detection of contrast material arrival in the aorta. Radiology 1997; 203:109-114.[Abstract/Free Full Text]
  8. Foo TKF, Saranathan M, Prince MR, Chenevert TL. Automated detection of bolus arrival and initiation of data acquisition in fast, three-dimensional, gadolinium-enhanced MR angiography. Radiology 1997; 203:275-280.[Abstract/Free Full Text]
  9. Prince MR, Grist TM, Debatin JF, eds. Basic concepts In: 3D contrast MR angiography. Berlin, Germany: Springer-Verlag, 1999; 13.
  10. Earls JP, Rofsky NM, DeCorato DR, Krinsky GA, Weinreb JC. Hepatic arterial-phase dynamic gadolinium-enhanced MR imaging: optimization with a test examination and a power injector. Radiology 1997; 202:268-273.[Abstract/Free Full Text]
  11. Lee VS, Lavelle MT, Rofsky NM, et al. Hepatic MR imaging with a dynamic contrast-enhanced isotropic volumetric interpolated breath-hold examination: feasibility, reproducibility and technical quality. Radiology 2000; 215:365-372.[Abstract/Free Full Text]
  12. Frederick MG, McElaney BL, Singer A, et al. Timing of parenchymal enhancement on dual-phase dynamic helical CT of the liver: how long does the hepatic arterial phase predominate? AJR Am J Roentgenol 1996; 166:1305-1310.[Abstract/Free Full Text]
  13. Hales MR, Allan JS, Hall EM. Injection-corrosion studies of normal and cirrhotic livers. Am J Pathol 1959; 5:909-941.
  14. Hoevels J, Lunderquist A, Tylen U. Spontaneous intermittent reversal of blood flow in intrahepatic portal vein branches in cirrhosis of the liver. Cardiovasc Radiol 1979; 2:267-273.[Medline]
  15. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:307-310.[CrossRef][Medline]
  16. Maki JH, Prince MR, Londy FJ, Chenevert TL. The effects of time varying intravascular signal intensity and k-space acquisition order on three-dimensional MR angiography image quality. J Magn Reson Imaging 1996; 6:642-651.[Medline]
  17. Murakami T, Kim T, Takamura M, et al. Hypervascular hepatocellular carcinoma: detection with double arterial-phase multi-detector row helical CT. Radiology 2001; 218:763-767.[Abstract/Free Full Text]
  18. Foley WD, Mallisee TA, Hohenwaltewr MD, Wilson CR, Quiroz FA, Taylor AJ. Multi-phase hepatic CT with multirow detector CT scanning. AJR Am J Roentgenol 2000; 175:679-685.[Abstract/Free Full Text]
  19. Larson RE, Semelka RC, Bagley AS, Molina PL, Brown ED, Lee JKT. Hypervascular malignant liver lesions: comparison of various MR imaging pulse sequences and dynamic CT. Radiology 1994; 192:393-399.[Abstract/Free Full Text]
  20. Paulson EK, Fisher AJ, DeLong DM, Parker DD, Nelson RC. Helical liver CT with computer-assisted bolus-tracking technology: is it possible to predict which patients will not achieve a threshold of enhancement? Radiology 1998; 209:787-792.[Abstract/Free Full Text]
  21. Mitsuzaki K, Yamashita Y, Ogata I, Tang Y, Namimoto T, Takahashi M. Optimal protocol for injection of contrast material at MR angiography: study of healthy volunteers. Radiology 1999; 213:913-918.[Abstract/Free Full Text]
  22. Tublin ME, Tessler FN, Peters TL, McGovern PC. Effect of injection rate of contrast medium on pancreatic and hepatic helical CT. Radiology 1999; 210:97-101.[Abstract/Free Full Text]
  23. Bae KT, Heiken JP, Brink JA. Aortic and hepatic contrast medium enhancement at CT. II. Effect of reduced cardiac output in a porcine model. Radiology 1998; 207:657-662.
  24. Garcia PA, Bonaldi VM, Bret PM, Liang L, Reinhold C, Atri M. Effect of rate of contrast medium injection on hepatic enhancement at CT. Radiology 1996; 199:185-189.[Abstract/Free Full Text]
  25. Sheiman RG, Raptopoulos V, Caruso P, Vrachliotis T, Pearlman J. Comparison of tailored and empiric scan delays for CT angiography in the abdomen. AJR Am J Roentgenol 1996; 167:725-729.[Abstract/Free Full Text]
  26. Materne R, Horsmans Y, Jamart J, Smith AM, Gigot JF, Van Beers BE. Gadolinium-enhanced arterial-phase MR imaging of hypervascular liver tumors: comparison between tailored and fixed scanning delays in the same patient. J Magn Reson Imaging 2000; 11:244-249.[CrossRef][Medline]
  27. Bendtsen F, Henriksen JH. Transit-time analysis in hepatic cirrhosis. Lancet 1999; 354:598-599.
  28. Soyer P, Dufresne AC, Somveille E, Scherrer A. MR imaging of the liver: effect of portal hypertension on hepatic parenchymal enhancement using a gadolinium chelate. J Magn Reson Imaging 1997; 7:142-146.[Medline]
  29. Van Beers BE, Materne R, Lacrosse M, et al. MR imaging of hypervascular liver tumors: timing optimization during the arterial-phase. J Magn Reson Imaging 1999; 9:562-567.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 RadiologyHome page
J. M. Willatt, H. K. Hussain, S. Adusumilli, and J. A. Marrero
MR Imaging of Hepatocellular Carcinoma in the Cirrhotic Liver: Challenges and Controversies
Radiology, May 1, 2008; 247(2): 311 - 330.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
L. L. Chu, B. N. Joe, A. C. A. Westphalen, E. M. Webb, F. V. Coakley, and B. M. Yeh
Patient-specific Time to Peak Abdominal Organ Enhancement Varies with Time to Peak Aortic Enhancement at MR Imaging
Radiology, December 1, 2007; 245(3): 779 - 787.
[Abstract] [Full Text] [PDF]

Home page
 RadioGraphicsHome page
S. N. Gandhi, M. A. Brown, J. G. Wong, D. A. Aguirre, and C. B. Sirlin
MR Contrast Agents for Liver Imaging: What, When, How
RadioGraphics, November 1, 2006; 26(6): 1621 - 1636.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hussain, H. K.
Right arrow Articles by Chenevert, T. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hussain, H. K.
Right arrow Articles by Chenevert, T. L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE