Published online before print January 19, 2006, 10.1148/radiol.2382041527
(Radiology 2006;238:920-928.)
© RSNA, 2006
Chronic Pancreatitis: MRCP versus ERCP for Quantitative Caliber Measurement and Qualitative Evaluation1
Ryo Tamura, MD,
Tadashi Ishibashi, MD and
Shoki Takahashi, MD
1 From the Department of Radiology, Northern Fukushima Medical Center, 23-1 Higashi, Hakozaki, Date, Fukushima 960-0502, Japan (R.T.); and Department of Radiology, Tohoku University School of Medicine, Sendai, Japan (T.I., S.T.). Received September 3, 2004; revision requested November 10; revision received January 4, 2005; accepted February 16; final version accepted April 24.
Address correspondence to R.T. (e-mail: tamura{at}jinsenkai.or.jp).
 |
ABSTRACT
|
|---|
Purpose: To retrospectively compare—in patients with chronic pancreatitis—magnetic resonance cholangiopancreatography (MRCP) and endoscopic retrograde cholangiopancreatography (ERCP) for measurement of main pancreatic duct (MPD) diameter by using area intensity measurement (AIM) at MRCP and full width at half maximum (FWHM) at ERCP and to retrospectively determine the accuracy of MRCP for depiction of pathologic changes by using ERCP as the reference standard.
Materials and Methods: The institutional review board approved this study and waived the need to obtain informed consent. Both MRCP and ERCP were performed in 24 patients with chronic pancreatitis (21 men, three women; mean age, 54 years ± 14 [standard deviation]). The diameter of the MPD was determined by using both methods at the same sites in the head, body, and tail of the pancreas. MRCP and ERCP measurements involved AIM and FWHM techniques, respectively. For qualitative evaluation, visualization of the MPD and pathologic findings was also examined by using both methods. Paired t and Wilcoxon matched-pair signed rank tests were performed for the quantitative and qualitative evaluations, respectively.
Results: The mean diameter of the MPD at ERCP was 1.5 times larger, on average, than that at MRCP; differences were statistically significant for each segment, as well as for the entire duct system. For qualitative evaluation, MRCP tended to be superior to ERCP for delineation of the MPD. Overall sensitivity, specificity, and accuracy values of MRCP for delineating pathologic pancreatic changes were 88% (87 of 99), 98% (44 of 45), and 91% (131 of 144), respectively.
Conclusion: Use of ERCP tends to result in overestimation of the caliber of the MPD. MRCP can enable accurate evaluation of the condition of the pancreatic duct and its changes in patients with chronic pancreatitis.
© RSNA, 2006
|
INTRODUCTION
|
|---|
Imaging is important in the diagnosis of chronic pancreatitis. In advanced chronic pancreatitis, dilatation of the main pancreatic duct (MPD) is an important pathologic finding. In chronic pancreatitis with mild pathologic changes, accurate imaging of irregularities of branch pancreatic ducts is necessary for a diagnosis (1). Dilatation of the MPD and irregularities of branch pancreatic ducts have been evaluated with endoscopic retrograde cholangiopancreatography (ERCP) as the accepted standard, but, because of major and relatively recent advances in both software and hardware, magnetic resonance cholangiopancreatography (MRCP) is often used for diagnosis. However, the spatial resolution of MRCP is still inferior to that of ERCP and thus is sometimes inadequate for evaluation of either duct diameter or duct wall irregularities in the pancreaticobiliary duct system. To our knowledge, results of no quantitative comparison between duct lumen diameters measured by using MRCP and those measured by using ERCP have been published.
In determining duct lumen diameter for quantitative analysis, full width at half maximum (FWHM) and full width at tenth maximum have been widely used. With these methods, a measurement is made by counting the pixels in digital images; hence, the accuracy of the measurement can be dependent on the pixel size. Currently, for clinical purposes, a pixel size of about 1 x 1 mm or larger is used for MRCP, and spatial resolution in the slab thickness direction is more than 1 mm (2–9). Because the normal upper limits of the diameter of the MPD are considered to be 4, 3, and 2 mm in the head, body, and tail of the pancreas, respectively (10), measurements acquired with the FWHM or the full width at tenth maximum method at MRCP performed with current techniques are inappropriate because of the inadequate spatial resolution yielded by such techniques. That may be why use of neither the FWHM nor the full width at tenth maximum method for this purpose has been described in the literature (to our knowledge) but performance of diameter measurements by using electronic calipers on the screen display or a similar approach has been described in the literature (9,11–14).
A new caliber or lumen diameter measurement method called area intensity measurement (AIM) has more recently been proposed (15). With this method, the lumen diameter is calculated on the basis of signal intensity on MRCP images, permitting a more accurate measurement of diameter than is possible by using FWHM. Thus, the purpose of our study was to retrospectively compare—in patients with chronic pancreatitis—MRCP and ERCP for measurement of the MPD diameter by using AIM at MRCP and FWHM at ERCP and to retrospectively determine the accuracy of MRCP for depiction of pathologic changes by using ERCP as the reference standard.
|
MATERIALS AND METHODS
|
|---|
Patient Population
This study involved 24 consecutive patients with chronic pancreatitis (21 men and three women) who ranged in age from 19 to 74 years (mean age, 54 years ± 14 [standard deviation]). These patients had undergone both MRCP and ERCP (after providing informed consent) between May 1997 and May 2000 at Tohoku University Hospital, Sendai, Japan, with a 1-month interval between the two procedures. We had institutional review board approval and waiver of informed consent for this retrospective study. The final diagnosis of chronic pancreatitis was reached after the possibility that a patient had a pancreatic neoplasm was excluded and with reference to clinical history, laboratory findings, and results of radiologic modalities such as ERCP, computed tomography, and ultrasonography (US). Twenty-one patients were considered to have alcoholic pancreatitis; one, autoimmune pancreatitis; and two, idiopathic pancreatitis.
Imaging Techniques
MRCP was performed by using a 1.5-T MR imaging unit (Magnetom Vision; Siemens, Erlangen, Germany) with a phased-array coil. Two series of MRCP images along the coronal or coronal oblique plane (–30° to +30° off the axis) were obtained. One series was obtained with a single-section method, and one series was obtained with a multiple-section method. The single-section method involved use of a full-Fourier-acquisition single-shot turbo spin-echo sequence (repetition time msec/echo time msec, 
/1100; flip angle, 150°; echo train length, 240; section thickness, 30–50 mm; number of sections, one; and acquisition time, 7 seconds); the multiple-section method involved use of a half-Fourier-acquisition single-shot turbo spin-echo sequence (/95; flip angle, 150°; echo train length, 128; section thickness, 4 mm; number of sections, 13; and acquisition time, 20 seconds). For each sequence, the field of view was 250 x 250 mm, the matrix was 240 x 256, and chemical fat suppression and oral contrast material (ferric ammonium citrate, FerriSeltz; Otsuka, Tokushima, Japan) were used. T1-weighted breath-hold spoiled gradient-echo fast low-angle shot images with fat suppression (100–150/4.0–4.1; flip angle, 70°; section thickness, 5 mm; intersection gap, 1 mm; number of sections, 15–20; field of view, 300–350 mm; and matrix, 256 x 144) were also acquired in the transverse plane.
Image Analysis
Qualitative analyses were performed in retrospective fashion by two diagnostic radiologists (R.T. and T.I., with 8 and 10 years of experience, respectively, in MRCP and ERCP image analysis) at a display monitor for MRCP and by using film for ERCP. The radiologists worked independently and had no knowledge of clinical data or other images. Both kinds of MRCP images (ie, those obtained with the single-section and those obtained with the multiple-section method) were evaluated. Final assessments were rendered in consensus. Quantitative analyses were performed at a monitor by one radiologist (R.T.).
Quantitative Analysis
The diameter of the MPD was measured in the head, body, and tail of the pancreas. Both MRCP and ERCP images were observed together so that the viewers could choose approximately corresponding sites near the center of the three segments along the long axes of the pancreatic duct. An overlap with branch ducts was avoided in measurement as much as was possible (Fig 1).
View larger version (109K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a: (a) Coronal single-section MRCP image (/1100) with fat suppression and (b) frontal ERCP image in 69-year-old man with chronic pancreatitis. Duct caliber measurements were performed at approximately the same sites at MRCP and ERCP. These sites were chosen near the center along the long axes of the head, body, and tail of the pancreas. An overlap with branch ducts was avoided during this measurement. Solid lines indicate measurement sites; dashed lines indicate borders of pancreatic segments.
|
|
View larger version (134K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b: (a) Coronal single-section MRCP image (/1100) with fat suppression and (b) frontal ERCP image in 69-year-old man with chronic pancreatitis. Duct caliber measurements were performed at approximately the same sites at MRCP and ERCP. These sites were chosen near the center along the long axes of the head, body, and tail of the pancreas. An overlap with branch ducts was avoided during this measurement. Solid lines indicate measurement sites; dashed lines indicate borders of pancreatic segments.
|
|
Measurement at MRCP
The measurement obtained by using MRCP was based on the AIM method, in which a caliber diameter is calculated on the basis of signal intensity (Appendix) (15). The procedure was performed as previously described. Briefly, a phantom containing tubes with various lumen diameters that were filled with distilled water was initially imaged 30 times for calibration with the same MRCP sequences used to image the patients. Scanned images were transported in Digital Imaging and Communications in Medicine, or DICOM, format to a personal computer (Power Macintosh G3; Apple, Cupertino, Calif), and the signal intensity curves of the lumens were obtained along the cross sections perpendicular to the ducts by using an image analysis application (NIH Image, version 1.62; National Institutes of Health, Washington, DC). Net area under the curve (AUC) values were calculated by subtracting the signal intensity of the pancreatic parenchyma (used as a background signal intensity value) from the gross signal intensity curves. Because all of the ducts were distinguishable from the background signal, no data were dropped from the analysis of results of MRCP of the phantom. The following regression equation was derived from the 30 images of the phantom and used to describe the relationship between the AUC value and the diameter of the duct: 
= (a · 
AUC) + b, where is the diameter of the duct as calculated in millimeters, the constant a is 0.22 ± 0.01 (standard deviation), and the constant b is 0.48 ± 0.11. The correlation coefficient between AUC and the diameter of the MPD was r = 0.967 (P < .001).
Each patient's MRCP data were transformed into DICOM format and analyzed at the personal computer by using the same method used in the phantom study. Finally, the diameter of the MPD in the three portions of the pancreas was calculated for each patient by using the regression equation and inserting the values of the AUC into the equation. MPD segments in which the signal intensity curve could not be distinguished from background signal were not measured.
Measurement at ERCP
At ERCP, the diameter of the MPD lumen was derived from the FWHM value, which was obtained from a curve that showed the profile of the duct. An image in an anteroposterior view with the best-expanded depiction of the MPD was selected from a series of ERCP images obtained in each patient. The film was digitized by scanning it with an image scanner (GT9000; Epson, Nagano, Japan). The data were transferred in tagged image file format (spatial resolution, 144 dots per inch; pixel size, 0.18 x 0.18 mm) to the personal computer. By using the same NIH image analysis application used with the MRCP images, MPD signal intensity curves were obtained at the sites described above and the FWHM was measured. Radiographic magnification was corrected for by performing a calibration measurement of the endoscope with its known diameter.
MPD diameters were compared among the head, body, and tail segments for each modality. The MPD diameter at MRCP was compared with that at ERCP for the three segments.
Qualitative Analysis
Qualitative visualization of MPD.—Qualitative visualization of the MPD at both MRCP and ERCP was evaluated. Visualization of the MPD was evaluated with a four-point scale according to the ratio of the visualized length to the full length of the MPD. With this scale, a grade of 1 indicated that this ratio was between 0% and 24%; a grade of 2, that the ratio was between 25% and 49%; a grade of 3, that the ratio was between 50% and 74%; and a grade of 4, that the ratio was between 75% and 100%. In determining the full length of the three segments of the MPD, we substituted the length of their corresponding pancreatic segments. The length of each of these segments was estimated on a composite image of the entire pancreas that was created from consecutive transverse images.
The visualization grades for the MPD were compared among the head, body, and tail segments for each modality. The grades at MRCP were also compared with those at ERCP for the three segments.
Qualitative visualization of pathologic findings.—Qualitative visualization and evaluation of the pancreas were performed at both MRCP and ERCP according to the criteria in the Cambridge classification of chronic pancreatitis (1). The pancreas was examined for the presence or absence of (a) irregularity in the MPD, (b) dilatation of branch ducts, (c) irregularity in branch ducts, (d) cystic lesion(s), (e) stenosis or obstruction of the MPD, and (f) pancreatic stone(s). For items a, d, e, and f, which did not involve branch ducts, the question was whether each condition was present or not. For items b and c, which concerned branch ducts, the results were considered positive only when three or more ducts were involved because the Cambridge classification criteria for chronic pancreatitis requires the involvement of at least three branch ducts before a pancreas is considered to show branch duct dilatation or irregularity. The ability to detect these conditions at MRCP was compared with the ability to detect them at ERCP, with ERCP considered to be the accepted reference standard. Cambridge classification of the condition of the patients in this series was attempted with ERCP.
Statistical Analysis
Sensitivity, specificity, and accuracy for detection of the pathologic findings were calculated. A two-tailed paired t test was performed for the quantitative study, and a Wilcoxon matched-pair signed rank test was performed for the qualitative study. Statistical tests and calculation of the regression equation in the AIM method were performed by using statistical software (StatView, version 5.0; Abacus Concepts, Berkeley, Calif). The relationship between results of qualitative visualization of the MPD at MRCP and those at ERCP was analyzed by using a generalized estimating equation adjustment for the clustering of segments within patients and by using statistical software (SAS, version 9.10; SAS Institute, Cary, NC). A P value of less than .05 was considered to indicate a statistically significant difference.
|
RESULTS
|
|---|
Quantitative Analysis
MRCP.—Sixty-seven (93%) of 72 MPD segments, including 24 segments in the head of the pancreas, 23 segments in the body of the pancreas, and 20 segments in the tail of the pancreas, were imaged clearly enough to be measured. The remaining five segments (one in the body and four in the tail of the pancreas) were excluded from the measurements because their signal intensity curves were not clear. The mean duct diameter was 3.5 mm ± 1.6 (standard deviation) in the head, 3.8 mm ± 1.9 in the body, and 3.3 mm ± 1.8 in the tail of the pancreas (overall mean, 3.6 mm ± 1.7) (Table 1). There was no significant difference in mean MPD diameter among the three segments.
ERCP.—Sixty-two (86%) of 72 MPD segments, including 23 segments in the head of the pancreas, 22 segments in the body of the pancreas, and 17 segments in the tail of the pancreas, were imaged clearly enough to be measured. The other 10 segments (one in the head, two in the body, and seven in the tail of the pancreas) did not sufficiently fill with contrast material and could not be measured. The mean duct diameter was 4.7 mm ± 2.2 in the head, 5.2 mm ± 2.8 in the body, and 4.3 mm ± 2.4 in the tail of the pancreas (overall mean, 4.7 mm ± 2.5) (Table 1). As at MRCP, at ERCP, there was no significant difference in mean MPD diameter among the three segments.
MRCP versus ERCP.—Of the 60 MPD segments that could be measured at both MRCP and ERCP, 50 (83%) appeared thicker at ERCP than at MRCP (Fig 2). The measured MPD diameter at ERCP was 1.5 times larger, on average, than that at MRCP (1.6 times larger in the head, 1.5 times larger in the body, and 1.3 times larger in the tail of the pancreas). The difference between the value at ERCP and that at MRCP was statistically significant (P < .01) for each segment as well as for the entire duct system.
View larger version (42K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2: Graph shows results of MPD caliber measurement at MRCP and ERCP. Of 60 segments of the MPD, 50 (83%) measured thicker at ERCP than at MRCP. The measured value at ERCP was 1.5 times thicker, on average, than that at MRCP, and the difference was significant (P < .001).
|
|
Qualitative Analysis
Qualitative visualization of MPD.—Although visualization grades at MRCP tended to be higher than grades at ERCP for each segment, the difference was not significant (Table 2). When we compared visualization grades between the segments, we found that the MPD tended to be imaged more poorly in the tail of the pancreas than in the head or the body at both MRCP and ERCP (Fig 3). According to results of both Wilcoxon matched-pair signed rank testing and generalized estimating equation analysis, this difference in visualization was significant at ERCP but not at MRCP.
View larger version (90K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3a: (a) Paracoronal single-section MRCP image (/1100) with fat suppression and (b) frontal ERCP image in 47-year-old woman with chronic pancreatitis. Despite the presence of a pancreatic duct stone (arrow), MRCP with a single-section sequence delineates the entire MPD. With ERCP, the MPD is not seen in the body and the tail of the pancreas owing to obstruction by the pancreatic duct stone.
|
|
View larger version (89K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3b: (a) Paracoronal single-section MRCP image (/1100) with fat suppression and (b) frontal ERCP image in 47-year-old woman with chronic pancreatitis. Despite the presence of a pancreatic duct stone (arrow), MRCP with a single-section sequence delineates the entire MPD. With ERCP, the MPD is not seen in the body and the tail of the pancreas owing to obstruction by the pancreatic duct stone.
|
|
Qualitative visualization of pathologic findings.—With ERCP considered as the accepted standard of reference, the overall sensitivity, specificity, and accuracy for depiction of pathologic changes with MRCP were 88% (87 of 99), 98% (44 of 45) and 91% (131 of 144), respectively, when these values were calculated with inclusion of all of the evaluated conditions (Table 3). When these values were calculated for individual conditions, MRCP was found to have the lowest sensitivity (76% [16 of 21]) for irregularity of branch ducts. According to results of Cambridge classification at ERCP, six (25%) of the 24 patients had mild pathologic changes, three (12%) had moderate pathologic changes, and 14 (58%) had marked pathologic changes indicative of chronic pancreatitis (Table 4).
View this table:
[in this window]
[in a new window]
|
Table 3. Results of Qualitative Visualization of Pathologic Findings at MRCP as Compared with Those at ERCP, the Reference-Standard Examination
|
|
|
DISCUSSION
|
|---|
Caliber Measurements with MRCP
In previous studies, the size of the MPD was quantitatively analyzed with MRCP (9,11–14,16,17). However, in these studies, lumen diameters were measured with electronic calipers or were presented without mention of what measurement methods were used, and the values were reported by using units of less than 1 pixel in size. Because the minimum measurement unit in digital imaging is pixel size, any reported values that included units less than 1 pixel in size would have been inaccurate, even if electronic calipers had been used. Specifically, for example, a measured value of 2.5 mm may not have been accurate if the pixel size, and the limit of resolution, was 1 mm.
Although there have been studies of quantitative analysis with MRCP, to our knowledge no investigators have actually reported results of a quantitative comparison of the caliber of the pancreatic duct at ERCP and MRCP. Investigators have stated that the duct appeared to be larger at ERCP than at MRCP (4–6). To our knowledge, ours is the first study in which MPD diameters measured at both ERCP and MRCP in the same patients were quantitatively compared. In the present study we measured the MPD in patients with chronic pancreatitis by using the AIM method of caliber measurement described in our previous report (15). In that original in vitro study, mean measurement errors for 1–8-mm caliber ducts were significantly smaller with the AIM method (mean measurement error, 6.8%) than with the FWHM method (mean measurement error, 30%). The AIM method is based on the supposition that a linear correlation between the volume of fluid and the signal intensity exists. The linear correlation and high reproducibility of this measurement have been confirmed in both previous studies with phantoms (15,18) and the present study with a phantom. Therefore, we believe that the MPD diameters measured at MRCP by using the AIM method in this study are more accurate than those measured with electronic calipers.
Quantitative Analysis and Discrepancies between MRCP and ERCP
In this study, 50 (83%) of 60 examined MPD segments had larger measurements at ERCP than at MRCP; the diameter was approximately 1.5 times larger at ERCP than at MRCP. This discrepancy between the two modalities is in agreement with suggestions in previous reports of MRCP (4–6). Three factors probably contributed to this discrepancy: First, the duct diameter may have been overestimated at ERCP owing to radiographic magnification. Second, the diameter may have been underestimated at MRCP owing to signal loss and blurring. Third, the duct was filled with a contrast medium for imaging with ERCP, and the injected contrast medium may have distended the duct (3,5,6,8,19). Radiographic magnification with ERCP was corrected for in this study, and measurements were made by obtaining the FWHM. Although an underestimation of duct size at MRCP could result from blurring and a lack of resolution, the data presented here suggest that the measurement discrepancies were primarily caused by overestimation at ERCP owing to the distending effect of the injected contrast medium on the pancreatic duct. In contrast, MRCP reveals duct size in more physiologic conditions. Bastid et al (20) compared the MPD diameter in chronic calcifying pancreatitis measured at US with that measured at ERCP. The mean diameters at US and ERCP were 4.2 and 5.4 mm, respectively. This discrepancy between ERCP and US appears to be similar to that reported here between ERCP and MRCP and provides additional support for the role of the distending effect of the contrast medium on the pancreatic duct at ERCP.
Qualitative Visualization of MPD
The diagnostic potential of MRCP in patients with pancreatitis has not been sufficiently analyzed (16). Previous studies in which MRCP and ERCP were directly and qualitatively compared were limited to small patient populations (3,4,16). The system used for qualitative grading of the MPD in this study is different from those used in the previous studies and was based only on the imaged length of the MPD for the purpose of objectivity. In this study, visibility grades at MRCP were slightly (but nonsignificantly) higher than grades at ERCP: At MRCP, the head, body, and tail segments of the MPD, respectively, had a visibility grade of 4 in 96%, 92%, and 83% of cases (mean, 90%); a grade of 1 in no, no, and 12% of cases (mean, 4%); and a grade of 2 in no, 4%, and 4% of cases (mean, 3%). These results were far superior to those in the study of Takehara et al (3). Although the grading system used was different, Takehara et al qualitatively compared MRCP and ERCP results and reported good visibility scores (meaning that the entire duct in each segment could be visualized) for 79%, 64%, and 53% of cases and poor visibility scores for 8%, 14%, and 17% of cases in the head, body, and tail segments, respectively. The better results achieved in our study may be related to recent advances in software and hardware and the fact that a larger proportion of patients in the present study had advanced pancreatic disease with dilated pancreatic ducts.
It has been reported that contrast resolution at imaging of the MPD in the tail of the pancreas is unsatisfactory when compared with resolution at imaging of the MPD in the head and body regions (21–23). Although the results of our study showed a similar tendency, the grades of the visualized lengths at MRCP were not significantly different among the three segments. At ERCP, however, the mean grade in the tail region was significantly lower than the mean grade in the head and body regions. This may be due to the fact that many patients in this study had advanced chronic pancreatitis with pancreatic stones and/or stenosis or extreme dilatation of the MPD. These factors might have made filling of the MPD with contrast medium at ERCP difficult, especially in the tail region. Although the difference was not statistically significant, the grades of the duct at MRCP were generally higher than those at ERCP, suggesting that MRCP is superior to ERCP in the delineation of the MPD, especially in advanced chronic pancreatitis.
Qualitative Visualization of Pathologic Findings
Previously, some authors have reported on the qualitative visualization of pathologic findings at MRCP and ERCP in patients with chronic pancreatitis. Takehara et al (3) reported that agreement between MRCP and ERCP was 83%–92% in cases of ductal dilation, 70%–92% in cases of ductal narrowing, and 92%–100% in cases of filling defects. Sica et al (24) reported that the specificity of MRCP in MPD characterization ranged from 86% to 88%. In the present study, overall sensitivity, specificity, and accuracy values for detecting lesions at MRCP were 88% (87 of 99), 98% (44 of 45), and 91% (131 of 144), respectively. These values are slightly superior to those in the previous studies.
In this study, among the pathologic conditions of the MPD, MRCP showed the highest specificity (100% [three of three]) but the lowest sensitivity (76% [16 of 21]) for depiction of irregularity in branch ducts. The high specificity with MRCP could be accounted for by the patient population in this study: More than half of all patients (58% [14 of 24]) had marked changes indicative of chronic pancreatitis according to the Cambridge classification. If our population had included more patients with less advanced pathologic changes, the scores might have been lower than those determined here. Pancreatic branch ducts narrower than 1 mm in diameter are not usually seen at MRCP (25). Therefore, these minute structures are difficult to evaluate not only quantitatively but also qualitatively. MRCP imaging of wall irregularities in branch ducts would require a higher spatial resolution and a higher signal-to-noise ratio. Thus, MRCP can depict relatively advanced changes well but might be inadequate for depicting subtle findings as compared with ERCP; this inadequacy should have been reflected in its high specificity but lower sensitivity for chronic pancreatitis.
Limitations
A potential limitation of this study was that the MRCP images were acquired with technology that is 5 or more years old.
In a small percentage of the segments (17% [10 of 60]) observed for the quantitative analysis in this study, the values for MPD lumen diameter were larger at MRCP than at ERCP. One possible reason for this may have been that there was an overlap of the MPD with branch ducts or with fluid in the gastrointestinal tract at MRCP, even though attempts were made to avoid this. Alternatively, the use of MRCP for separate phantom studies and at different times for clinical examination might have led to a slight variability in signal. However, we can find no clear explanation for this inconsistency.
For qualitative visualization of the pancreas in terms of pathologic findings, evaluation for each patient was performed for the entire MPD but not for individual duct segments. Use of this method of evaluation—looking at the duct as a whole—with both procedures might have provided a less precise groundwork for the comparison of detectable lesions between MRCP and ERCP.
In summary, we compared MRCP and ERCP both qualitatively and quantitatively in patients with chronic pancreatitis. We used the AIM method for measurements with MRCP, and the duct lumen diameter was calculated on the basis of signal intensity, while FWHM was used with ERCP. The average diameter of the MPD at ERCP was 1.5 times larger than that at MRCP; the difference was statistically significant. For visual or qualitative evaluation, MRCP tended to be superior to ERCP in the delineation of the MPD and provided satisfactorily high diagnostic accuracy in the evaluation of pathologic findings.
|
APPENDIX
|
|---|
The AIM is a caliber measurement method that is based on the presumption that the sum of the signal intensities of an object in a slab is proportional to the volume of the object (15). For this method to be valid, the entire duct lumen should be in the slab and its content must consist of uniform material.
In our previous study (15), a platelike phantom containing tubes of various lumen diameters (1, 2, 2.5, 3, 3.5, 4, 5, 6, 7, and 8 mm) that were filled with distilled water and arrayed parallel to each other was scanned with a MRCP sequence in a direction parallel to the plate of the phantom (Fig A1).
View larger version (108K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure A1: MRCP image of phantom obtained with single-section sequence (/1100). The phantom consisted of tubes with various lumen diameters (1, 2, 2.5, 3, 3.5, 4, 5, 6, 7, and 8 mm). (Reprinted, with permission, from reference 15.)
|
|
Then, signal intensity curves for the tube lumens were obtained (Fig A2) from a cross section perpendicular to the tubes (Fig A1). Net AUCs for the individual tubes were obtained from the curve (shaded portions in Fig A2) and increased with increasing diameter. There was a high correlation between tube diameter and AUC (Fig A3).
View larger version (36K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure A2: Signal intensity curve for 1–8-mm-caliber ducts obtained from cross section perpendicular to duct axis (white vertical line in Fig A1). Shaded portions represent the AUC values for individual tubes. (Reprinted, with permission, from reference 15.)
|
|
Once a regression equation between lumen diameter and the AUC was derived from the phantom imaging data, the diameter of an examined duct (eg, the MPD in a patient) could be calculated by inserting the AUC value obtained from a signal intensity curve into the equation. The measurement error associated with the AIM was significantly (P < .001) lower than that associated with the FWHM measurement, which was also examined for comparison by using the same phantom.
|
ACKNOWLEDGMENTS
|
|---|
The authors acknowledge the contributions to this work from Tooru Shimosegawa, MD, PhD, and Masaru Koizumi, MD, PhD, of the Department of Gastroenterology; Ichiro Tsuji, MD, PhD, of the Division of Epidemiology of the Department of Public Health and Forensic Medicine; and Kei Takase, MD, PhD, Takayuki Yamada, MD, PhD, and Syuuichi Higano, MD, PhD, of the Department of Diagnostic Radiology of Tohoku University School of Medicine, Sendai, Japan.
|
FOOTNOTES
|
|---|
Abbreviations: AIM = area intensity measurement • AUC = area under the curve • ERCP = endoscopic retrograde cholangiopancreatography • FWHM = full width at half maximum • MPD = main pancreatic duct • MRCP = MR cholangiopancreatography
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, R.T., T.I., S.T.; study concepts/study design or data acquisition or data analysis/interpretation, R.T., T.I., S.T.; manuscript drafting or manuscript revision for important intellectual content, R.T., T.I., S.T.; approval of final version of submitted manuscript, R.T., T.I., S.T.; literature research, R.T., T.I., S.T.; clinical and experimental studies, R.T., T.I., S.T.; statistical analysis, R.T., T.I., S.T.; and manuscript editing, R.T., T.I., S.T.
|
References
|
|---|
- Axon AT, Classen M, Cotton PB, Cremer M, Freeny PC, Lees WR. Pancreatography in chronic pancreatitis: international definitions. Gut 1984;25(10):1107–1112.[Abstract/Free Full Text]
- Matos C, Deviere J, Cremer M, Nicaise N, Struyven J, Metens T. Acinar filling during secretin-stimulated MR pancreatography. AJR Am J Roentgenol 1998;171(1):165–169.[Abstract/Free Full Text]
- Takehara Y, Ichijo K, Tooyama N, et al. Breath-hold MR cholangiopancreatography with a long-echo-train fast spin-echo sequence and a surface coil in chronic pancreatitis. Radiology 1994;192(1):73–78.[Abstract/Free Full Text]
- Soto JA, Barish MA, Yucel EK, Ferrucci JT. MR cholangiopancreatography: findings on 3D fast spin-echo imaging. AJR Am J Roentgenol 1995;165(6):1397–1401.[Free Full Text]
- Barish MA, Yucel EK, Soto JA, Chuttani R, Ferrucci JT. MR cholangiopancreatography: efficacy of three-dimensional turbo spin-echo technique. AJR Am J Roentgenol 1995;165(2):295–300.[Abstract/Free Full Text]
- Reinhold C, Bret PM. Current status of MR cholangiopancreatography. AJR Am J Roentgenol 1996;166(6):1285–1295.[Abstract/Free Full Text]
- Bret PM, Reinhold C, Taourel P, Guibaud L, Atri M, Barkun AN. Pancreas divisum: evaluation with MR cholangiopancreatography. Radiology 1996;199(1):99–103.[Abstract/Free Full Text]
- Holzknecht N, Gauger J, Sackmann M, et al. Breath-hold MR cholangiography with snapshot techniques: prospective comparison with endoscopic retrograde cholangiography. Radiology 1998;206(3):657–664.[Abstract/Free Full Text]
- Manfredi R, Costamagna G, Brizi MG, et al. Pancreas divisum and "santorinicele": diagnosis with dynamic MR cholangiopancreatography with secretin stimulation. Radiology 2000;217(2):403–408.[Abstract/Free Full Text]
- Roberts-Thompson IC. Endoscopic retrograde pancreatography: analysis of the normal pancreatogram and changes which are associated with chronic pancreatitis and pancreatic cancer. Med J Aust 1977;2(24):793–796.[Medline]
- Matos C, Metens T, Deviere J, et al. Pancreatic duct: morphologic and functional evaluation with dynamic MR pancreatography after secretin stimulation. Radiology 1997;203(2):435–441.[Abstract/Free Full Text]
- Manfredi R, Costamagna G, Brizi MG, et al. Severe chronic pancreatitis versus suspected pancreatic disease: dynamic MR cholangiopancreatography after secretin stimulation. Radiology 2000;214(3):849–855.[Abstract/Free Full Text]
- Fukukura Y, Fujiyoshi F, Sasaki M, Nakajo M. Pancreatic duct: morphologic evaluation with MR cholangiopancreatography after secretin stimulation. Radiology 2002;222(3):674–680.[Abstract/Free Full Text]
- Cappeliez O, Delhaye M, Deviere J, et al. Chronic pancreatitis: evaluation of pancreatic exocrine function with MR pancreatography after secretin stimulation. Radiology 2000;215(2):358–364.[Abstract/Free Full Text]
- Tamura R, Ishibashi T, Saito H, et al. New duct caliber measurement methods on magnetic resonance cholangiopancreatography [in Japanese]. Nippon Igaku Hoshasen Gakkai Zasshi 2000;60(13):738–745.[Medline]
- Hellerhoff KJ, Helmberger H 3rd, Rosch T, Settles MR, Link TM, Rummeny EJ. Dynamic MR pancreatography after secretin administration: image quality and diagnostic accuracy. AJR Am J Roentgenol 2002;179(1):121–129.[Abstract/Free Full Text]
- Manfredi R, Lucidi V, Gui B, et al. Idiopathic chronic pancreatitis in children: MR cholangiopancreatography after secretin administration. Radiology 2002;224(3):675–682.[Abstract/Free Full Text]
- Heverhagen JT, Muller D, Battmann A, et al. MR hydrometry to assess exocrine function of the pancreas: initial results of noninvasive quantification of secretion. Radiology 2001;218(1):61–67.[Abstract/Free Full Text]
- Soto JA, Barish MA, Yucel EK, et al. Pancreatic duct: MR cholangiopancreatography with a three-dimensional fast spin-echo technique. Radiology 1995;196(2):459–464.[Abstract/Free Full Text]
- Bastid C, Sahel J, Filho M, Sarles H. Diameter of the main pancreatic duct in chronic calcifying pancreatitis: measurement by ultrasonography versus pancreatography. Pancreas 1990;5(5):524–527.[Medline]
- Fulcher AS, Turner MA, Capps GW, Zfass AM, Baker KM. Half-Fourier RARE MR cholangiopancreatography: experience in 300 subjects. Radiology 1998;207(1):21–32.[Abstract/Free Full Text]
- Miyazaki T, Yamashita Y, Tsuchigame T, Yamamoto H, Urata J, Takahashi M. MR cholangiopancreatography using HASTE (half-Fourier acquisition single-shot turbo spin-echo) sequences. AJR Am J Roentgenol 1996;166(6):1297–1303.[Abstract/Free Full Text]
- Irie H, Honda H, Tajima T, et al. Optimal MR cholangiopancreatographic sequence and its clinical application. Radiology 1998;206(2):379–387.[Abstract/Free Full Text]
- Sica GT, Braver J, Cooney MJ, Miller FH, Chai JL, Adams DF. Comparison of endoscopic retrograde cholangiopancreatography with MR cholangiopancreatography in patients with pancreatitis. Radiology 1999;210(3):605–610.[Abstract/Free Full Text]
- Fulcher AS, Turner MA. MR pancreatography: a useful tool for evaluating pancreatic disorders. RadioGraphics 1999;19(1):5–24.[Abstract/Free Full Text]
TOP
ABSTRACT
INTRODUCTION
· D + ß, where D is the lumen diameter and