HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2383041384v1 239/1/122    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Nasu, K. Articles by Ueda, T. Search for Related Content PubMed PubMed Citation Articles by Nasu, K. Articles by Ueda, T.

Hepatic Metastases: Diffusion-weighted Sensitivity-encoding versus SPIO-enhanced MR Imaging¹

¹ From the Department of Diagnostic Radiology, National Cancer Center Hospital East, 6-5-1 Kashiwanoha, Kashiwa, Chiba 277-8577, Japan (K.N., Y.K., S.N.); Department of Diagnostic Radiology, National Cancer Center Hospital, Tokyo, Japan (S.K.); Department of Radiology, Kofu Kyoritu Hospital, Yamanashi, Japan (T.T.); and Department of Radiology, Chiba University, School of Medicine, Chiba, Japan (S.Y., K.M., T.U.). From the 2003 RSNA Annual Meeting. Received August 11, 2004; revision requested October 21; revision received March 9, 2005; accepted April 4; final version accepted May 13. Address correspondence to K.N. (e-mail: kanasu{at}east.ncc.go.jp).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Purpose: To retrospectively compare accuracy of diffusion-weighted (DW) single-shot echo-planar imaging with sensitivity encoding (SENSE) with that of superparamagnetic iron oxide (SPIO)-enhanced magnetic resonance (MR) imaging in the evaluation of hepatic metastases due to extrahepatic malignancies.

Materials and Methods: Patients provided informed consent; ethics committee approval was not required. The data of 24 patients (16 men, eight women; age range, 41–68 years; mean age, 61.9 years) with 40 resected hepatic metastases were retrospectively reviewed. Before SPIO administration, DW SENSE imaging and T2-weighted fast spin-echo (SE) and T1-weighted dual-echo fast field-echo (FFE) MR imaging were performed. After SPIO administration, T2-weighted fast SE, T1-weighted dual-echo, and T2*-weighted FFE MR examinations were performed. Images were divided into two sets: The SPIO-enhanced MR image set consisted of pre- and postcontrast T2-weighted fast SE and dual-echo T1-weighted FFE images and postcontrast T2*-weighted FFE images. The DW SENSE image set included DW SENSE images and precontrast T2-weighted fast SE and dual-echo T1-weighted FFE images. Three radiologists individually interpreted these images and sorted the confidence levels for presence of hepatic metastasis in each section into five grades. Area under the receiver operating characteristic (ROC) curve (A_z) was calculated for each image set.

Results: Hepatic metastases showed higher signal intensity on DW SENSE images than on T2-weighted fast SE images. Conversely, signals from vessels and cysts were suppressed with DW SENSE imaging. ROC analysis showed higher A_z values when the DW SENSE image set was interpreted (0.90) than when the SPIO-enhanced MR image set was interpreted (0.81). The sensitivity and specificity, respectively, of total cases were 0.66 and 0.90, for the SPIO-enhanced MR image set and 0.82 and 0.94 for the DW SENSE image set. During SPIO-enhanced MR image interpretation, lesions 1 cm in diameter or smaller showed significantly lower sensitivity than lesions larger than 1 cm in diameter. During both interpretation sessions, left lobe lesions showed significantly lower sensitivity than right lobe lesions.

Conclusion: Combined reading of DW SENSE images and T2-weighted fast SE and dual-echo T1-weighted FFE MR images showed higher accuracy in the detection of hepatic metastases than did reading of SPIO-enhanced MR images.

© RSNA, 2006

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
Surgical resection is the most effective and curative treatment for hepatic metastases (1–4). It has been proved that hepatic resection with the aim of total removal of tumor tissue improves the prognosis in patients with metastatic liver disease deriving from colorectal cancer (2–4). Thus, surgeons are increasingly demanding accurate information from diagnostic radiologists that will enable them to determine the exact number and location of liver metastases. To provide this information, radiologists have used computed tomography (CT) during arterioportography (5,6) and superparamagnetic iron oxide (SPIO)-enhanced magnetic resonance (MR) imaging (7–9). SPIO-enhanced MR imaging has high sensitivity that matches that of CT during arterioportography and higher specificity than that of CT during arterioportography (8,9). The primary advantage of SPIO-enhanced MR imaging is that, unlike CT during arterioportography, it is noninvasive. Many doctors now regard SPIO-enhanced MR imaging as the best available examination technique in the evaluation of hepatic metastases.

On the other hand, diffusion-weighted (DW) single-shot echo-planar imaging is known for its high contrast resolution. It has been reported that this sequence depicts not only hyperintense fresh brain infarctions but also hyperintense malignant hepatic tumors (10,11). For the following reasons, however, clinical application of this sequence in abdominal imaging is still rare: First, the image acquisition technique that involves the use of echo-planar imaging is sensitive to susceptibility. As a result, the images become extremely distorted by the air in the lungs and the intestinal tract. Second, chemical shift artifacts cause severe misregistration of fat tissue, which makes it impossible to obtain reliable positional information in the abdomen, where fat tissue is more plentiful than in the head (12,13).

However, the rapid progress of parallel imaging techniques, especially sensitivity encoding (SENSE), is changing this situation. By making use of sensitivity differences among multiple surface coils, SENSE can be used to reduce the number of phase-encoding steps without loss of spatial resolution (14,15). This technique reduces acquisition time, minimizes echo-planar imaging artifacts, and improves the quality of images obtained with DW single-shot echo-planar imaging (16). Hereafter, DW single-shot echo-planar imaging with simultaneous use of SENSE will be referred to as DW SENSE imaging. The purpose of our study was to use receiver operating characteristic (ROC) analysis to retrospectively compare the accuracy of DW SENSE imaging with that of SPIO-enhanced MR imaging for evaluation of hepatic metastases (17).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
We reported our intended study to our ethics committee; however, the committee indicated that its approval was not required. At the time of imaging, patients provided consent, which allowed use of their data for research purposes.

Patient Group
Surgical resection of hepatic metastases was performed in 62 patients at the National Cancer Center Hospital East between August 2002 and January 2004. MR examinations, including DW SENSE imaging and SPIO-enhanced MR imaging, were performed in 24 patients before surgery. The patient group comprised the 24 patients with surgically resected and pathologically proved hepatic metastases. This group comprised 16 men and eight women (age range, 41–77 years; mean age, 61.9 years). The primary cancer sites in each patient were as follows: 20 colorectal cancers, one gastric cancer, one common bile duct cancer, one lung cancer, and one gastrointestinal stromal tumor of the jejunum. Each primary tumor generated up to four metastatic liver tumors, with one nodule generated in each of 13 patients; two nodules, in each of seven patients; three nodules, in each of three patients; and four nodules, in one patient. A total of 40 nodules were included in the study. Intraoperative ultrasonography revealed no other metastatic hepatic tumors in these patients. A pathologic examination revealed that these nodules were consistent with metastases generated by the primary neoplasm in each patient. Of the 40 nodules, eight were located in the left lobe and the rest were located in the right lobe. These nodules were measured on MR images and were 0.5–6.0 cm (mean, 1.9 cm) along the long axis; 18 were 1 cm or smaller in size. MR examinations were performed in every patient within 2 weeks before surgery.

Imaging Protocols
MR examinations were performed with a 1.5-T MR imager (Gyroscan Intera Master; Philips Medical Systems, Best, the Netherlands) and a SENSE body coil. These apparatuses and sequences were all commercially available and purchased from the same manufacturer as release 8.1 (Philips Medical Systems). The following imaging sequences were performed before SPIO administration:

DW SENSE imaging was performed with the following parameters: repetition time msec/echo time msec, 1600/73; b factors, 0 and 500 sec/mm²; spectral presaturation with inversion recovery for fat suppression; matrix size, 256 x 97; half scan factor, 0.693; reduction factor of SENSE, two; field of view, 35 x 28 cm; four signals acquired; section thickness, 7 mm; section gap, 1 mm; 22 transverse sections acquired; respiration trigger; and mean imaging time, 20 seconds.

The motion-probing gradient pulses of DW SENSE imaging were placed along the x-, y-, and z-axes. During image interpretation, we used only the trace images synthesized from the three images in which the motion-probing gradient pulses were placed in each direction.

T2-weighted fast spin-echo (SE) MR imaging was performed with the following parameters: 4062/90; echo train length, 15; spectral presaturation with inversion recovery for fat suppression; matrix, 512 x 211; reduction factor, two; field of view, 35 x 28 cm; one signal acquired; section thickness, 7 mm; section gap, 1 mm; 22 transverse sections acquired; end-expiration breath hold; and acquisition time, 26 seconds.

Dual-echo T1-weighted fast field-echo (FFE) MR imaging was performed with the following parameters: repetition time, 150 msec; opposed-phase echo time, 2.3 msec; in-phase echo time, 4.6 msec; flip angle, 80°; matrix, 512 x 211; reduction factor, two; field of view, 35 x 28; two signals acquired; section thickness, 7 mm; section gap, 1 mm; 22 transverse sections acquired; end-expiration breath hold; and acquisition time, 28 seconds.

After the previously mentioned sequences were performed, SPIO was administered intravenously. In 10 patients, ferumoxides (Feridex; Tanabe Seiyaku, Osaka, Japan) was administered at a dose of 0.015 mmol of iron per kilogram of body weight, diluted in 100 mL of 5% glucose, and infused over 20 minutes; imaging commenced 30 minutes after the start of intravenous infusion. In 14 patients, ferucarbotran (Resovist; Nihon Schering, Osaka, Japan) was injected as a rapid bolus and immediately followed by a saline solution flush, thus yielding a dose of 0.008 mmol of iron per kilogram of body weight; imaging commenced 10 minutes after injection. Subsequently, T2-weighted fast SE and dual-echo T1-weighted FFE MR imaging sequences were performed again. T2*-weighted FFE MR imaging also was performed, with the following parameters: 320/10; flip angle, 70°; matrix, 256/179; reduction factor, two; field of view, 35 x 28 cm; one signal acquired; section thickness, 7 mm; section gap, 1 mm; 22 transverse sections acquired; end-expiration breath hold; and acquisition time, 28 seconds.

All sequences covered the whole liver. The section thickness, section gap, and field of view were occasionally changed depending on the size of the liver. The only preparation before the examination was an 8-hour fasting period. During the image acquisition phase of the (a) T2-weighted fast SE, (b) dual-echo T1-weighted FFE, and (c) T2*-weighted FFE MR imaging sequences, oxygen inhalation was provided to make it easier for the patient to hold his or her breath for a longer period of time.

Imaging Analysis
Two radiologists from the National Cancer Center Hospital East (K.N. and Y.K., with 15 and 14 years of experience in abdominal MR imaging, respectively) retrospectively reviewed all images with knowledge of the results of surgical resection and confirmed the presence of pathologically proved hepatic metastases (n = 40) on MR images. From all obtained images, the anatomic section levels passing around the center of the metastatic hepatic lesions were selected in each sequence. Among the section levels, four each had two metastases. Thus, the number of section levels with positive signals was 36 (32 section levels each with one tumor and four section levels each with two tumors). A total of 54 section levels without metastases were randomly selected from the remaining section levels; thus, the number 54 has no intrinsic meaning itself. Consequently, 90 section levels were selected for analysis (ie, 36 + 54 = 90). Eight images were obtained at each section level: one DW SENSE image, two T2-weighted fast SE images (one precontrast, one postcontrast), four dual-echo T1-weighted FFE images (two precontrast, two postcontrast), and one postcontrast T2*-weighted FFE image. Finally, 720 images were chosen for ROC analysis (ie, 90 x 8 = 720). The same two radiologists interpreted all selected images again and confirmed that no other hepatic nodules were caused by suspected metastasis, with the exception of the previously mentioned 40 nodules. A total of 11 benign nodules, which were consistent with simple cysts, were also identified in the selected section levels. To match the anatomic section levels of selected images among different sequences, respiratory misregistration was corrected by the same two radiologists on the basis of mutual agreement.

The SPIO-enhanced MR image set and the DW SENSE image set were prepared from these images. The SPIO-enhanced MR image set consisted of (a) pre- and postcontrast T2-weighted fast SE, (b) pre- and postcontrast dual-echo T1-weighted FFE, and (c) postcontrast T2*-weighted FFE MR images. Seven images were obtained for each section level. The DW SENSE image set included precontrast T2-weighted fast SE, precontrast dual-echo T1-weighted FFE, and DW SENSE images. Four images were obtained for each section level. The images corresponding to each section level in each image set were displayed separately on one film hard copy. Thus, each image set consisted of 90 film hard copies.

Three trained diagnostic radiologists certified by the Japan Radiology Society (S.Y., K.M., and T.U., with 13, 12, and 11 years of experience in abdominal MR imaging, respectively) individually interpreted MR images without being provided any clinical information. They were not told how many hepatic lesions were in each section level. The image-reading sessions began with interpretation of the SPIO-enhanced MR imaging set. Four weeks later, these same radiologists reviewed the DW SENSE image set. The order of the section levels was randomized before every reading session to minimize the influence of memory.

The readers scored individual section levels for the presence or absence of metastatic hepatic lesions and assigned the following confidence levels to their observations: 1, definitely or almost definitely absent; 2, probably absent; 3, possibly present; 4, probably present; and 5, definitely or almost definitely present. The readers also referred to the segment in which they suspected the metastasis. If they determined that there were multiple lesions in one section level, each suspected nodule was scored.

For interpretation of the SPIO-enhanced MR image set, the criteria for identification of metastases were solid nodules showing more prominent signal intensity on postcontrast T2-weighted fast SE images than on precontrast T2-weighted fast SE images and satisfying either of the following conditions: (a) ring enhancement was observed around the nodule on any postcontrast dual-echo T1-weighted FFE image or (b) the nodules showed obviously higher signal intensity on postcontrast T2*-weighted FFE MR images than did the liver parenchyma (18,19). On the other hand, the criteria for identification of metastasis at interpretation of the DW SENSE image set were as follows: solid nodules that (a) could be identified on DW SENSE images and any of the T2-weighted fast SE images or dual-echo T1-weighted FFE MR images and (b) showed more prominent signal intensity on DW SENSE images than on T2-weighted fast SE MR images. These criteria were routinely used at the National Cancer Center Hospital East, so we were confident about their appropriateness in empirical situations. These criteria were presented as rough reference standards to the readers, however, and the final decision was left to the discretion of each reader.

For each image set, an ROC curve was fitted to each reader's confidence scoring. The diagnostic accuracy of each image set was determined by calculating the area under each reader-specific ROC curve (A_z). Computer software (Stata 8.0; Stata, College Station, Tex) was used to depict ROC curves and calculate A_z values. The total number of cases represented the results from the combined data of the three readers, and the composite ROC curve and mean A_z value were calculated with this combined data. A_z values for the two methods calculated by each reader and combined data were compared by using the nonparametric method described by Hanley and McNeil (20) and the previously mentioned computer software .

We also calculated sensitivity and specificity in each reading session for each reader. Nodules with confidence scores of 4 and 5 were considered positive, and nodules with confidence scores of 1–3 were considered negative. These criteria were given to the readers before reading sessions. In this study, more than one lesion was obtained from each patient. Thus, we used generalized estimating equations to revise the data clustering and dependency (21). In the generalized estimating equations, we assessed the relative fraction of sensitivity and specificity among the following factors: inter-method, inter-observer, tumor diameter (1 cm or >1 cm), and tumor location (left lobe or right lobe). In this estimation, the link function was set at log link, and an independent working correction matrix was used. We also assessed the relative fraction of sensitivity of each method among the following factors: inter-observer differences, tumor diameter, and tumor location. In this estimation, statistical models were constructed in each MR imaging method. Relative fractions of sensitivity were estimated with generalized estimating equations. The link function was set at log link, and an independent working correction matrix was used. The computer software used to apply the generalized estimating equations was SAS, version 8.02 (SAS Institute, Cary, NC); the Proc Genmod command was used.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 
DW SENSE imaging displayed somewhat similar contrast compared with that displayed with T2-weighted fast SE imaging. However, there were major differences between these two sequences with respect to the following factors: The metastatic tumors clearly demonstrated better signal intensity on DW SENSE images than on T2-weighted fast SE images. Hepatic cysts demonstrated various signal intensities on DW SENSE images, although they had the shared characteristic of demonstrating lower signal intensity on DW SENSE images than on T2-weighted fast SE images. The signal intensity of intrahepatic vessels was low on DW SENSE images, unlike that seen on T2-weighted fast SE MR images. The normal hepatic parenchyma had a tendency to show slightly higher signal intensity on DW SENSE images than on T2-weighted fast SE MR images. Accordingly, the metastatic hepatic lesions were always the most hyperintense nodular lesions on DW SENSE images (Fig 1); however, some imaging artifacts were evident. Above all, the most frequently encountered artifact was subcardiac signal loss in the left lobe. Abnormally hyperintense areas that were consistent with susceptibility artifacts were occasionally observed around air in the lungs and bowel tract; however, the image quality of DW SENSE images was generally good enough to allow interpretation. There were no image distortions or misregistrations of fat tissue.

View larger version (181K): [in this window] [in a new window] [Download PPT slide]   Figure 1a: (a) Precontrast T2-weighted fast SE MR image (4062/90) and (b) DW SENSE image (1600/73; b factor, 500 sec/mm2) obtained in a 59-year-old man with a hepatic metastasis (short arrow) caused by colorectal cancer in segment 7 and a small cyst (arrowhead) in segment 8. Metastatic tumor shows higher signal intensity on b than on a. Conversely, the cyst demonstrates lower signal intensity on b than on a. Intrahepatic vessels (long arrow) show low signal intensity on b, which is different from the signal intensity of the intrahepatic vessels seen on a.

View larger version (120K): [in this window] [in a new window] [Download PPT slide]   Figure 1b: (a) Precontrast T2-weighted fast SE MR image (4062/90) and (b) DW SENSE image (1600/73; b factor, 500 sec/mm2) obtained in a 59-year-old man with a hepatic metastasis (short arrow) caused by colorectal cancer in segment 7 and a small cyst (arrowhead) in segment 8. Metastatic tumor shows higher signal intensity on b than on a. Conversely, the cyst demonstrates lower signal intensity on b than on a. Intrahepatic vessels (long arrow) show low signal intensity on b, which is different from the signal intensity of the intrahepatic vessels seen on a.

View larger version (20K): [in this window] [in a new window] [Download PPT slide]   Figure 2: Composite ROC curves for three readers in each interpretation session show reader confidence in detection of hepatic metastases with DW SENSE imaging and SPIO-enhanced MR imaging. Mean Az values are 0.90 and 0.81, respectively. The difference between these indexes is significant (P < .001). FPF = false-positive fraction, TPF = true-positive fraction.

View larger version (166K): [in this window] [in a new window] [Download PPT slide]   Figure 3a: Images obtained in a 65-year-old man with a small liver metastasis caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 0.8-cm-diameter metastasis (arrow) is depicted in segment 5 on every image; however, it is difficult to indicate this lesion on a–e. All readers overlooked this nodule during SPIO-enhanced MR image interpretation. Conversely, this lesion can be easily identified on e. All readers could detect this nodule during DW SENSE image interpretation.

View larger version (185K): [in this window] [in a new window] [Download PPT slide]   Figure 3b: Images obtained in a 65-year-old man with a small liver metastasis caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 0.8-cm-diameter metastasis (arrow) is depicted in segment 5 on every image; however, it is difficult to indicate this lesion on a–e. All readers overlooked this nodule during SPIO-enhanced MR image interpretation. Conversely, this lesion can be easily identified on e. All readers could detect this nodule during DW SENSE image interpretation.

View larger version (179K): [in this window] [in a new window] [Download PPT slide]   Figure 3c: Images obtained in a 65-year-old man with a small liver metastasis caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 0.8-cm-diameter metastasis (arrow) is depicted in segment 5 on every image; however, it is difficult to indicate this lesion on a–e. All readers overlooked this nodule during SPIO-enhanced MR image interpretation. Conversely, this lesion can be easily identified on e. All readers could detect this nodule during DW SENSE image interpretation.

View larger version (155K): [in this window] [in a new window] [Download PPT slide]   Figure 3d: Images obtained in a 65-year-old man with a small liver metastasis caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 0.8-cm-diameter metastasis (arrow) is depicted in segment 5 on every image; however, it is difficult to indicate this lesion on a–e. All readers overlooked this nodule during SPIO-enhanced MR image interpretation. Conversely, this lesion can be easily identified on e. All readers could detect this nodule during DW SENSE image interpretation.

View larger version (169K): [in this window] [in a new window] [Download PPT slide]   Figure 3e: Images obtained in a 65-year-old man with a small liver metastasis caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 0.8-cm-diameter metastasis (arrow) is depicted in segment 5 on every image; however, it is difficult to indicate this lesion on a–e. All readers overlooked this nodule during SPIO-enhanced MR image interpretation. Conversely, this lesion can be easily identified on e. All readers could detect this nodule during DW SENSE image interpretation.

View larger version (162K): [in this window] [in a new window] [Download PPT slide]   Figure 3f: Images obtained in a 65-year-old man with a small liver metastasis caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 0.8-cm-diameter metastasis (arrow) is depicted in segment 5 on every image; however, it is difficult to indicate this lesion on a–e. All readers overlooked this nodule during SPIO-enhanced MR image interpretation. Conversely, this lesion can be easily identified on e. All readers could detect this nodule during DW SENSE image interpretation.

View larger version (152K): [in this window] [in a new window] [Download PPT slide]   Figure 4a: Images obtained in a 74-year-old woman with multiple liver metastases caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 2-cm-diameter metastasis (short arrow) is clearly seen in segment 8 on every image. Another small metastasis (long arrow) with a diameter smaller than 1 cm is shown in segment 3. This small metastasis is located in the subcardiac signal loss artifact (arrowheads in f) and is hardly detected. In the DW SENSE image reading session, all readers overlooked this small nodule.

View larger version (192K): [in this window] [in a new window] [Download PPT slide]   Figure 4b: Images obtained in a 74-year-old woman with multiple liver metastases caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 2-cm-diameter metastasis (short arrow) is clearly seen in segment 8 on every image. Another small metastasis (long arrow) with a diameter smaller than 1 cm is shown in segment 3. This small metastasis is located in the subcardiac signal loss artifact (arrowheads in f) and is hardly detected. In the DW SENSE image reading session, all readers overlooked this small nodule.

View larger version (175K): [in this window] [in a new window] [Download PPT slide]   Figure 4c: Images obtained in a 74-year-old woman with multiple liver metastases caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 2-cm-diameter metastasis (short arrow) is clearly seen in segment 8 on every image. Another small metastasis (long arrow) with a diameter smaller than 1 cm is shown in segment 3. This small metastasis is located in the subcardiac signal loss artifact (arrowheads in f) and is hardly detected. In the DW SENSE image reading session, all readers overlooked this small nodule.

View larger version (166K): [in this window] [in a new window] [Download PPT slide]   Figure 4d: Images obtained in a 74-year-old woman with multiple liver metastases caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 2-cm-diameter metastasis (short arrow) is clearly seen in segment 8 on every image. Another small metastasis (long arrow) with a diameter smaller than 1 cm is shown in segment 3. This small metastasis is located in the subcardiac signal loss artifact (arrowheads in f) and is hardly detected. In the DW SENSE image reading session, all readers overlooked this small nodule.

View larger version (176K): [in this window] [in a new window] [Download PPT slide]   Figure 4e: Images obtained in a 74-year-old woman with multiple liver metastases caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 2-cm-diameter metastasis (short arrow) is clearly seen in segment 8 on every image. Another small metastasis (long arrow) with a diameter smaller than 1 cm is shown in segment 3. This small metastasis is located in the subcardiac signal loss artifact (arrowheads in f) and is hardly detected. In the DW SENSE image reading session, all readers overlooked this small nodule.

View larger version (122K): [in this window] [in a new window] [Download PPT slide]   Figure 4f: Images obtained in a 74-year-old woman with multiple liver metastases caused by colorectal cancer. Precontrast (a) T2-weighted fast SE (4062/90), (b) dual-echo T1-weighted FFE opposed-phase (150/2.3, 80° flip angle), and (c) dual-echo T1-weighted FFE in-phase (150/4.6, 80° flip angle) MR images. Postcontrast (d) T2-weighted fast SE (4062/90) and (e) T2*-weighted FFE (320/10, 70° flip angle) MR images. (f) DW SENSE image (1600/73; b factor, 500 sec/mm2). A 2-cm-diameter metastasis (short arrow) is clearly seen in segment 8 on every image. Another small metastasis (long arrow) with a diameter smaller than 1 cm is shown in segment 3. This small metastasis is located in the subcardiac signal loss artifact (arrowheads in f) and is hardly detected. In the DW SENSE image reading session, all readers overlooked this small nodule.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

Surgeons ask diagnostic radiologists not to simply pinpoint metastases that are difficult to identify on conventional MR images but to identify all metastases, however small. Thus, the previously mentioned drawbacks to SPIO-enhanced MR imaging are serious. In response, many radiologists have tried to resolve these problems by improving the imaging sequences used in SPIO-enhanced MR imaging (18,19,23). Although the sequences used in this study were optimized as much as possible given past reports and our own experience, the results of this study show that the deficits in SPIO-enhanced MR imaging have not been overcome.

To our knowledge, the usefulness of DW imaging in the abdomen has not been established; however, its potential has been suggested in some reports (10,11). For example, Kim et al (11) reported that malignant hepatic tumors, especially hepatic metastases, have a lower apparent diffusion coefficient than benign nodules. DW imaging has the potential to be useful; however, it is rarely used in the abdomen because of the poor quality of DW images in prior studies.

The basic issues of DW imaging and the current development of parallel imaging techniques need to be verified. The method used to obtain a DW signal was established early in the MR imaging development process (24). Motion-probing gradient pulses placed before and after a 180° radiofrequency pulse for SE generate an echo influence by means of diffusion. If the whole k-space is filled with this DW echo, actual image acquisition takes too long; however, a long echo train is produced by echo-planar imaging immediately after generation of the DW echo that rapidly fills the k-space. This method is called DW single-shot echo-planar imaging, and it is the most commonly used technique in the clinical context (25); however, this imaging technique has some important defects. The images are often accompanied by intense distortion and misregistration of fat tissue. Parallel imaging techniques, including SENSE, are dramatically improving this situation. With the use of SENSE, the later half of the echo train—which includes much influence of susceptibility and chemical shift—is no longer needed to produce images. This is why the image quality of DW SENSE imaging is superior to that of previously described examinations.

When DW SENSE imaging was used, the metastatic hepatic tumors showed apparently higher signal intensity and the cysts demonstrated lower signal intensity than that seen with T2-weighted fast SE MR imaging. We believe this phenomenon is caused by the following factors: The Brownian motion of extracellular water molecules around the cancer cells is restricted because of the high cellularity of each mass, while the fluids in the cysts consist of free water molecules, which yield a higher apparent diffusion coefficient (26,27). However, DW imaging is influenced by both diffusion and T2 relaxation time. Thus, tissues with a long T2 relaxation time have a tendency to appear hyperintense on DW images, even if they have a large apparent diffusion coefficient (28). The T2 shine-through effect is assumed to be the reason the cysts showed various signal intensities. In every patient, the cysts showed lower signal intensity with DW SENSE imaging than with T2-weighted fast SE MR imaging, allowing most of the cysts to be easily differentiated from metastatic tissues. On the other hand, suppression of intrahepatic vascular signal is caused by different mechanisms. In DW imaging, only the protons—on which a couple of motion-probing gradient pulses were placed—can emit signals. Thus, signals from rapidly moving protons, such as those in blood vessels, were completely suppressed. Accordingly, it was easy to differentiate metastases from tangential sections of thin vessels with DW SENSE imaging.

As mentioned previously, the problems with SPIO-enhanced MR imaging were almost resolved with DW SENSE imaging. It was a natural result for the A_z value of the DW SENSE image set to be higher than that of the SPIO-enhanced MR image set. In particular, there was no doubt about the usefulness of DW SENSE imaging in the identification of hepatic nodules measuring 1 cm or smaller. Metastatic tumors tended to appear larger on DW images than on T2-weighted fast SE images. The reason for this phenomenon is unclear, although it is assumed to contribute to the high sensitivity for detection of small metastases at DW SENSE imaging. As mentioned previously, however, the need for combined interpretation of DW SENSE images and T2-weighted fast SE MR images must be emphasized. In this study, apparent diffusion coefficient measurement was not performed because it fell outside of the aim of this study; however, we think that the visual evaluation of signal intensity change between these two sequences relies on intuitive interpretation of the apparent diffusion coefficient. The decision criteria in the DW SENSE image interpretation session were created on the basis of our understanding, as described previously.

This study also revealed several shortcomings of DW SENSE imaging. The artifacts caused by physiologic movement were a serious problem. In particular, the subcardiac signal loss, which appeared most frequently, significantly decreased sensitivity for detection of metastases in the left lobe. This artifact was assumed to be caused by the same mechanism of intrahepatic vascular suppression. More specifically, cardiac motion in the systolic phase shook the adjacent liver within the duration of motion-probing gradient pulses, which caused attenuation of the signal from the left hepatic lobe. Simultaneous use of cardiac triggering with respiratory gating that would enable two complete motion-probing gradient pulses within the diastolic phase during expiration may be one way to resolve this problem; however, this procedure takes too much time for it to be performed in patients at this time. These are the reasons we acquired four signals with DW SENSE imaging. With an increased number of signals acquired, the probability of data acquisition during the diastolic phase increases. Of course, this is not an ideal solution; however, we believe that this method is more practical than simultaneous use of cardiac and respiratory triggering.

This study has limitations. The majority of metastases were located in the right lobe, and the selected sections did not include hepatic hemangiomas. The image selection in this study was not intentional, but the bias may decrease the reliability of this study. However, hepatic metastases are generated more frequently in the right lobe than in the left lobe. Thus, DW SENSE imaging is dependable in the detection of hepatic metastases. We know that hepatic hemangioma, which is another common benign hepatic nodule, showed more various and complicated patterns on DW SENSE images than did cysts. These matters must be assessed in future studies.

In conclusion, combined image interpretation with DW SENSE imaging and T2-weighted fast SE and dual-echo T1-weighted FFE MR imaging yielded better accuracy in the detection of hepatic metastases than did SPIO-enhanced MR imaging; these examinations have the potential to replace SPIO-enhanced MR imaging in the preoperative evaluation of hepatic metastases in the near future.

	ACKNOWLEDGMENTS

 
We are grateful to our colleagues—Yoshito Kato, Narumi Akimoto, and Hiroyuki Fujikawa—for their contribution in image acquisition. We also thank Makoto Obara and Marc van Cauteren for their useful suggestion regarding our MR apparatus and software. We express gratitude for the assistance of Yoshitaka Murakami regarding statistical analysis of this study.

	FOOTNOTES

 

Abbreviations: A_z = area under the ROC curve • DW = diffusion weighted • FFE = fast field echo • ROC = receiver operating characteristic • SE = spin echo • SENSE = sensitivity encoding • SPIO = superparamagnetic iron oxide

Author contributions: Guarantors of integrity of entire study, K.N., S.N.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, K.N., S.K.; clinical studies, K.N., S.K., T.T.; statistical analysis, K.N.; and manuscript editing, K.N.

Authors stated no financial relationship to disclose.

	References

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION References

 

1. Okano K, Maeba T, Ishimura K, et al. Hepatic resection for metastatic tumors from gastric cancer. Ann Surg 2002;235:86–91.[CrossRef][Medline]
2. Penna C, Nordlinger B. Colorectal metastasis (liver and lung). Surg Clin North Am 2002;82:1075–1090.[CrossRef][Medline]
3. Yamada H, Katoh H, Kondo S, Okushiba S, Morikawa T. Repeat hepatectomy for recurrent hepatic metastases from colorectal cancer. Hepatogastroenterology 2001;48:828–830.[Medline]
4. Fong Y, Salo J. Surgical therapy of hepatic colorectal metastasis. Semin Oncol 1999;26:514–523.[Medline]
5. Soyer P, Levesque M, Caudron C, Elias D, Zeitoun G, Roche A. MRI of liver metastases from colorectal cancer vs CT during arterial portography. J Comput Assist Tomogr 1993;17:67–74.[Medline]
6. Schmidt J, Strotzer M, Fraunhofer S, Boedeker H, Zirngibl H. Intraoperative ultrasonography versus helical computed tomography and computed tomography with arterioportography in diagnosing colorectal liver metastases: lesion-by-lesion analysis. World J Surg 2000;24:43–47.[CrossRef][Medline]
7. Seneterre E, Taourel P, Bouvier Y, et al. Detection of hepatic metastases: ferumoxides-enhanced MR imaging versus unenhanced MR imaging and CT during arterial portography. Radiology 1996;200:785–792.[Abstract/Free Full Text]
8. Vogl TJ, Schwarz W, Blume S, et al. Preoperative evaluation of malignant liver tumors: comparison of unenhanced and SPIO (Resovist)-enhanced MR imaging with biphasic CTAP and intraoperative US. Eur Radiol 2003;13:262–272.[Medline]
9. Strotzer M, Gmeinwieswer J, Schmidt J, et al. Diagnosis of liver metastasis from colorectal adenocarcinoma: comparison of spiral-CTAP combined with intravenous contrast-enhanced spiral-CT and SPIO-enhanced MR combined with plain MR imaging. Acta Radiol 1997;38:986–992.[Medline]
10. Ichikawa T, Haradome H, Hachiya J, Nitatori T, Araki T. Diffusion-weighted MR imaging with a single-shot echoplanar sequence: detection and characterization of focal hepatic lesions. AJR Am J Roentgenol 1998;170:397–402.[Abstract/Free Full Text]
11. Kim T, Murakami T, Takahashi S, Hori M, Tsuda K, Nakamura H. Diffusion-weighted single-shot echoplanar MR imaging for liver disease. AJR Am J Roentgenol 1999;173:393–398.[Abstract/Free Full Text]
12. Schmithorst VJ, Dardzinaki BJ, Holland SK. Simultaneous correlation of ghost and geometric distortion artifacts in EPI using a multiecho reference scan. IEEE Trans Med Imaging 2001;20:535–539.[CrossRef][Medline]
13. Schick F, Forster J, Machann J, Kuntz R, Claussen CD. Improved clinical echo-planar MRI using spatial-spectral excitation. J Magn Reson Imaging 1998;8:960–967.[Medline]
14. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: sensitivity-encoding for fast MRI. Magn Reson Med 1999;42:952–962.[CrossRef][Medline]
15. Kurihara Y, Yakushiji YK, Tani I, van Cauteren M. Coil sensitivity-encoding in MR imaging: advantage and disadvantage in clinical practice. AJR Am J Roentgenol 2002;178:1087–1091.[Free Full Text]
16. Bammer R, Keeling SL, Augustin M, et al. Improved diffusion-weighed single-shot echo-planar imaging (EPI) in stroke using sensitivity-encoding (SENSE). Magn Reson Med 2001;46:548–554.[CrossRef][Medline]
17. Metz CE. ROC methodology in radiologic imaging. Invest Radiol 1986;21:720–733.[Medline]
18. Poeckler-Schoeniger C, Koepke J, Gueckel F, Sturm J, Georgi M. MRI with superparamagnetic iron oxide: efficacy in the detection and characterization of focal hepatic lesions. Magn Reson Imaging 1999;17:383–392.[CrossRef][Medline]
19. Kim JH, Kim MJ, Suh SH, Chung JJ, Yoo HS, Lee JT. Characterization of focal hepatic lesions with ferumoxides-enhanced MR imaging: utility of T1-weighted spoiled gradient recalled echo images using different echo times. J Magn Reson Imaging 2002;15:573–583.[CrossRef][Medline]
20. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36.[Abstract/Free Full Text]
21. Pepe MS. The statistical evaluation of medical tests for classification and prediction: Oxford statistical science series 28. New York, NY: Oxford University Press, 2003; 35–65.
22. Ward J, Feng C, Guthrie JA, et al. Hepatic lesion detection after superparamagnetic iron oxide enhancement: comparison of five T2-weighted sequences at 1.0 T by using alternative-free response receiver operating characteristic analysis. Radiology 2000;214:159–166.[Abstract/Free Full Text]
23. Kim SH, Choi D, Lim JH. Optimal pulse sequence for ferumoxides-enhanced MR imaging used in the detection of hepatocellular carcinoma: a comparative study using seven pulse sequences. Korean J Radiol 2002;3:87–97.[Medline]
24. Stejskal EO, Tanner JE. Spin echoes in the presence of a time-dependent field gradient. J Chem Phys 1965;42:288–292.[CrossRef]
25. Baker LL, Kucharczyk J, Sevick RJ, Mintorovitch J, Moseley ME. Recent advances in MR imaging/spectroscopy of cerebral ischemia. AJR Am J Roentgenol 1991;156:1133–1143.[Abstract/Free Full Text]
26. Sugahara T, Korogi Y, Kochi M. Usefulness of diffusion-weighted MRI with echo-planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 1999;9:53–60.[CrossRef][Medline]
27. Gauvain KM, McKinstry RC, Mukherjee P, et al. Evaluating pediatric brain tumor cellularity with diffusion-tensor imaging. AJR Am J Roentgenol 2001;177:449–454.[Abstract/Free Full Text]
28. Burdette JH, Elster AD, Ricci PE. Acute cerebral infarction: quantification of spin-density and T2 shine-through phenomenon on diffusion-weighted MR images. Radiology 1999;212:333–339.[Abstract/Free Full Text]

This article has been cited by other articles:

				H. Kandpal, R. Sharma, K. S. Madhusudhan, and K. S. Kapoor Respiratory-Triggered Versus Breath-Hold Diffusion-Weighted MRI of Liver Lesions: Comparison of Image Quality and Apparent Diffusion Coefficient Values Am. J. Roentgenol., April 1, 2009; 192(4): 915 - 922. [Abstract] [Full Text] [PDF]

				A. C. Silva, J. M. Evans, A. E. McCullough, M. A. Jatoi, H. E. Vargas, and A. K. Hara MR Imaging of Hypervascular Liver Masses: A Review of Current Techniques RadioGraphics, March 1, 2009; 29(2): 385 - 402. [Abstract] [Full Text] [PDF]

				W. Luboldt, R. Kufer, N. Blumstein, T. L. Toussaint, A. Kluge, M. D. Seemann, and H.-J. Luboldt Prostate Carcinoma: Diffusion-weighted Imaging as Potential Alternative to Conventional MR and 11C-Choline PET/CT for Detection of Bone Metastases Radiology, December 1, 2008; 249(3): 1017 - 1025. [Abstract] [Full Text] [PDF]

				M. Bruegel, J. Gaa, S. Waldt, K. Woertler, K. Holzapfel, B. Kiefer, and E. J. Rummeny Diagnosis of Hepatic Metastasis: Comparison of Respiration-Triggered Diffusion-Weighted Echo-Planar MRI and Five T2-Weighted Turbo Spin-Echo Sequences Am. J. Roentgenol., November 1, 2008; 191(5): 1421 - 1429. [Abstract] [Full Text] [PDF]

				Y. Ohno, H. Koyama, Y. Onishi, D. Takenaka, M. Nogami, T. Yoshikawa, S. Matsumoto, Y. Kotani, and K. Sugimura Non-Small Cell Lung Cancer: Whole-Body MR Examination for M-Stage Assessment--Utility for Whole-Body Diffusion-weighted Imaging Compared with Integrated FDG PET/CT Radiology, August 1, 2008; 248(2): 643 - 654. [Abstract] [Full Text] [PDF]

				T. Parikh, S. J. Drew, V. S. Lee, S. Wong, E. M. Hecht, J. S. Babb, and B. Taouli Focal Liver Lesion Detection and Characterization with Diffusion-weighted MR Imaging: Comparison with Standard Breath-hold T2-weighted Imaging Radiology, March 1, 2008; 246(3): 812 - 822. [Abstract] [Full Text] [PDF]

				D.-M. Koh and D. J. Collins Diffusion-Weighted MRI in the Body: Applications and Challenges in Oncology Am. J. Roentgenol., June 1, 2007; 188(6): 1622 - 1635. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2383041384v1 239/1/122    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Nasu, K. Articles by Ueda, T. Search for Related Content PubMed PubMed Citation Articles by Nasu, K. Articles by Ueda, T.