- Email: [email protected]
Does endoscopic digital pancreatography constitute an advance in pancreatic imaging?
Does endoscopic digital pancreatography constitute an advance in pancreatic imaging? Generoso Uomo, MD Rosaria de Ritis, MD Pier Giorgio Rabitti, MD Alfonso Ragozzino, MD
Background: From a technological viewpoint, there have been few advances in retrograde pancreatography since the inception of ERCP despite remarkable advances in other radiographic techniques such as CT and magnetic resonance imaging. We describe a new method of pancreatographic imaging, endoscopic digital pancreatography, based on digital image intensification. Methods: Thirty-two nonconsecutive patients with various pancreatic diseases were entered into the study. Pancreatography was performed using standard technique with a non-ionic, low osmolarity contrast medium. In addition to conventional pancreatographic x-rays, digital images were acquired and then processed by means of computer-assisted technical modalities. Both pancreatograms were evaluated by two independent investigators. Results: Digital pancreatography was rated as superior to conventional pancreatography in terms of overall quality as well as specific parameters. Xray dosage was significantly higher and fluoroscopy time significantly longer for conventional pancreatography. Conclusion: Endoscopic digital pancreatography represents a significant advance, one with several important advantages compared with conventional pancreatography. Endoscopic retrograde pancreatography (ERP) has, since its introduction almost 30 years ago, had a major impact on the assessment of patients with pancreatic disorders.1-5 Despite the extensive application of ERP, both diagnostic and therapeutic, there have been relatively few novel changes in the procedure from a technologic point of view,4,5 espe-
Received August 30, 1996. For revision August 19, 1997. Accepted March 11, 1998. From the Pancreas Unit, Department of Gastroenterology and Radiology, Cardarelli Hospital, Naples, Italy. Reprint requests: G. Uomo, MD, Pancreas Unit, Department of Gastroenterology, Cardarelli Hospital, via Cardarelli 9, 80131 Napoli, Italy. Copyright © 1998 by the American Society for Gastrointestinal Endoscopy 0016-5107/98/$5.00 + 0 37/69/90175 VOLUME 48, NO. 1, 1998
cially by comparison with the evolution of other radiologic techniques for the study of the pancreas, such as CT and magnetic resonance. This report describes preliminary results with a new pancreatographic imaging method based on digital image intensification that which we call endoscopic digital pancreatography (EDP). X-ray exposure was also measured. In our study, digital pancreatograms were rated as superior in overall quality and a variety of specific parameters compared with conventional pancreatograms. PATIENTS AND METHODS Thirty-two nonconsecutive patients were entered into the study over an 18-month period (July 1994 to December 1995). Cannulation of the main duodenal papilla was achieved in 30 patients. There were 13 men and 17 women; mean age was 56 ± 15.6 years (range 24 to 77 years). The indications for visualization of the pancreatic duct (PD) were as follows: suspected chronic pancreatitis (16 patients), suspected pancreatic cancer (9 patients), idiopathic acute pancreatitis (3 patients), and pseudocyst post-acute pancreatitis (2 patients). Duodenoscopy was performed by one operator (G.U.) using standard technique and equipment.4 Non-ionic, low osmolarity contrast medium (iopamidol) was used in all cases. Digital acquisition of images was initiated when early PD injection was evident on fluoroscopy; total acquisition was usually completed in a matter of seconds and never required longer than thirty seconds. EDP was performed using a radioscopic/graphic unit (Siregraph C; Siemens, Erlangen, Germany) on line with a computer (Digitron 3VAI; Siemens Medical Service, Chicago, Ill.). Images were achieved in continuous mode with a rate of four frames per second (120 images in 30 seconds). The automatic exposure control features included fluoroscopically programmed radiography in which the correct parameters for each exposure were automatically calculated and set from the fluoroscopic parameters. The digital imaging chain provided an image (512 × 512 matrix size) with a 10-bit gray scale, series exposures, and automatic image optimization. Conventional ERP was also performed for each patient during the same session using the same radiographic unit. After EDP completion, images were processed by means of computer-assisted technical modalities including dynamic zoom, magnification, image-subtraction, automatic and manual adjustment of gray-scale, calibration parameters, and video-tape recording. Moreover, a special filter (Sobel, Digitron) was used when an enhancement of PD edges was considered to be helpful. All of these noted computer-assisted processing methods required just a few seconds: for example, filtration required 10 seconds. Each pancreatogram obtained with EDP and ERP was evaluated by two skilled independent observers who rated the overall quality of the images, as well as the view of the lumen and edges of the main pancreatic duct together with its secondary and tertiary branches. The images GASTROINTESTINAL ENDOSCOPY
67
G Uomo, R de Ritis, P Rabitti, et al.
Endoscopic digital pancreatography: An advance in pancreatic imaging?
Table 1. Numbers of patients with either poor-sufficient or excellent image ratings for overall quality, and lumen and edges of the main pancreatic duct, secondary and tertiary branches. Poor-sufficient
Excellent
5 15
25 (p = 0.006) 15
Overall quality EDP ERP Edge Poor-sufficient MPD EDP ERP Secondary branches EDP ERP Tertiary branches EDP ERP
Lumen Excellent
Poor-sufficient
Excellent
2 14
28 (p = 0.01) 16
2 10
28 (p = 0.01) 20
9 13
21 17
9 14
21 (p = 0.01) 16
9 12
21 (p = 0.01) 18
19 20
11 10
EDP, Endoscopic digital pancreatography; ERP, conventional retrograde pancreatography; MPD, main pancreatic duct.
were judged as poor-sufficient or excellent. When the evaluations of the two observers did not agree (four cases), a joint revaluation was carried out. X-ray exposure was measured in an additional six EDP and six ERP procedures not included in the study. Dosage was measured by means of termoluminescent dosimeters placed at the level of the thyroid and left eyelid of the endoscopist nearest to the x-ray table and of the patient. Fluoroscopy time was also recorded for both types of examination. Statistical analysis was performed using Student’s ttest, Fisher exact test, and Friedman test, as indicated. Significance was set at p < 0.05.
RESULTS Numbers of patients with either poor-sufficient or excellent ratings of EDP and ERP images with regard to overall quality and the lumen and edges of the main pancreatic duct together with its secondary and tertiary branch ducts are shown in Table 1. Examples of EDP images and corresponding standard ERP images are presented in Figures 1 through 3. EDP was especially useful in patients with pancreatic cancer. In seven patients with a confirmed diagnosis of malignancy, there were significant differences in the ratings for EDP versus ERP images in terms of overall quality, main pancreatic duct edge, and the edge and lumen of secondary ducts. However, differences in the several rating parameters did not reach statistical significance in the group of 13 patients with confirmed chronic pancreatitis. X-ray dosage was significantly higher for ERP compared with EDP at the level of eyelid both for patients (10.4 ± 2.7 mGy in ERP, 2.4 ± 0.9 mGy in EDP; p < 0.001) and operator (0.3 ± 0.2 mGy in ERP, 68
GASTROINTESTINAL ENDOSCOPY
0.07 ± 0.02 in EDP; p = 0.01). Fluoroscopy time was significantly longer for ERP (120 ± 21.5 seconds) than EDP (24 ± 4.2 seconds); p < 0.001). DISCUSSION Digital image intensifier radiology, a sophisticated technique that uses an x-ray image intensifier and a television system with a digitized video signal, is still undergoing relatively rapid development.6 It has been used in recent years for a broad range of examinations including barium-contrast studies of the gastrointestinal tract and several iodine-based contrast studies.6-8 Digital technique offers many advantages in comparison with conventional radiology.6-9 These include (1) an increase in image quality due to acquisition of a real-time series of images and various types of analysis of images (dynamic high-resolution fluoroscopic images, single-shot and series exposures on the monitor, images recorded on film); (2) post-processing image-handling capabilities (adjustment of gray-scale value and contrast, edge enhancement, magnification) that provide for optimal image presentation; (3) digital storage of images that is easier and less expensive than conventional methods of image archiving and also allows for image transfer in a communication network; and, (4) reduced radiation exposure because less radiation is required per exposure and there are fewer unsatisfactory exposures. Moreover, digital radiology requires far fewer standard x-ray films with a corresponding reduction in the environmental load associated with processing of chemical and silver halides. The major disadvantage of digital VOLUME 48, NO. 1, 1998
Endoscopic digital pancreatography: An advance in pancreatic imaging?
Figure 1. Cancer of the body of the pancreas confirmed at surgery. A, Conventional pancreatography demonstrates interruption and stenosis of the main pancreatic duct at the junction of the head and body (small arrows) and an irregular filling defect in the accessory duct (large arrow). B, EDP visualization in the same patient of the main pancreatic duct at level of body and tail also demonstrates irregular, longstenosis and better resolution of the secondary and tertiary branches in the head of the gland as well as better resolution of the stenosis in the accessory pancreatic duct. VOLUME 48, NO. 1, 1998
G Uomo, R de Ritis, P Rabitti, et al.
Figure 2. Cancer in the head of the gland that was confirmed at surgery. A, Conventional pancreatogram: the presence of contrast medium in the duodenum covers a suspected stenosis of the main pancreatic duct. B, EDP of same patient with subtraction of duodenal contents and use of an edge enhancement filter provides clear visualization of an irregular stenosis (arrows) of the main pancreatic duct.
GASTROINTESTINAL ENDOSCOPY
69
G Uomo, R de Ritis, P Rabitti, et al.
Endoscopic digital pancreatography: An advance in pancreatic imaging?
Figure 3. EDP with subtraction of osseous and gaseous interferences shows parenchymography of the tail of the gland. Surgical exploration revealed a pancreatic adenocarcinoma.
imaging relates to its limited spatial resolution which renders it less suitable for large input screens, such as in chest examination.6 The application of digital radiologic technique to pancreatography has been reported by Bickston et al.10 These investigators performed digital pancreatography in 10 patients and judged the images obtained to be superior to those achievable with conventional radiologic technique. In the present study, the availability of high-quality, real-time images with automatic optimization and post-processing modalities significantly improved the overall quality of pancreatographic images and displayed the main pancreatic duct together with its secondary and tertiary branches in greater detail. Through use of the dynamic acquisition capability of EDP we were able to monitor the progressive filling and opacification of the pancreatic ducts. This proved to be a crucial factor in the precise identification of communications between the ductal system and parenchymal cavities and also made it easier to recognize segmental ductal overfilling. Moreover, because of enhanced contrast and resolution, EDP recognized impending acinarization earlier than was possible by ERP (Fig. 3). Based on this last observation and the lower concentrations of contrast medium generally used in digital technique10 and considering that acinarization is an important risk factor in pancreatic inflammation associated with retrograde pancreatography,11 it is possible that EDP might reduce the frequency of procedure-related pancreatitis compared with ERP. 70
GASTROINTESTINAL ENDOSCOPY
In effect, EDP provides the endoscopist with more precise control of the injection phase of the procedure. The limited spatial resolution of EDP in comparison with ERP was evident during the acquisition phase of the procedure. However, a satisfactory result can be obtained by means of the post-processing modalities mentioned previously. Digital recording of images makes it possible to review the entire examination minute by minute as compared with the study of single x-ray pictures as is the case with conventional ERP. This can be especially advantageous when dealing with an evolving lesion such as a pseudocyst. From the economic standpoint, EDP costs around $200 per examination which is more or less in line with our cost for a traditional ERP. However, this is the operational cost and does not take into account the fact that a digital computerized radiology unit was already available in our hospital and does not include the costs for technical personnel and physicians fees. It is often necessary to provide referring physicians with “hard copy” images of pancreatographic images. With EDP, there is less need for duplicate images. By selecting images, it is possible to demonstrate the most significant findings with fewer images, which becomes a cost-saving factor. EDP significantly reduced x-ray exposure for both patients and operators when compared with ERP. In addition to a significant reduction in fluoroscopy time, there were fewer attempts to better demonstrate a finding by repeated injections. In addition to their high quality, the images recorded after filling is complete can be reviewed repeatedly without the need for further x-ray exposures. In conclusion, our experience and data suggest that EDP should be considered a true advance in pancreatic imaging, one that offers several important advantages in comparison with conventional pancreatography. Studies of larger series of patients are needed to fully investigate the potential advantages and limitations of this new technique including cost-effectiveness, potential to reduce procedurerelated complications, and enhancement of pancreatographic diagnosis. REFERENCES 1. Axon ATR, Classen M, Cotton PB. Pancreatography in chronic pancreatitis: international definitions. Gut 1984;25:1107-12. 2. Geenen JE, ASGE distinguished lecture. Endoscopic therapy of pancreatic disease: a new horizon. Gastrointest Endosc 1988;34:386-91. 3. Schapiro RH. ERCP in the diagnosis of pancreatic and biliary disease. In: Jacobson IM, eds. ERCP, diagnostic and therapeutic applications. Elseiver, New York, 1989;9-39. 4. Siegel JH, editor. Endoscopic retrograde cholangiopancreVOLUME 48, NO. 1, 1998
atography. Technique, diagnosis and therapy. New York: Raven Press; 1992. 5. Huibregtse K, Smits ME. Endoscopic management of disease of the pancreas. Am J Gastroenterol 1994;89:S67-77. 6. Lehmann K, Busch HP, Georgi M. Digital fluoroscopic radiology: evaluation of clinical impact. Eur J Radiol 1993;17:3-7. 7. Gingold EL, Tucker DM, Barnes GT. Computed radiography: user-programmable features and capabilities. J Digital Imag 1994;7:113-8.
VOLUME 48, NO. 1, 1998
8. Busch HP, Georgi M. Digital radiographs. Quality assurance and radiation protection. Radiol Diagn 1992;33:321-4. 9. Torrent M, Moreau M, Deschamps D, Capron J. Digital imaging in conventional radiology. Medica Mundi 1995;40:3-11. 10. Bickston SJ, Balfe DM, Brink JA, Aliperti G. Digital pancreatography is superior to conventional pancreatography [abstract]. Gastrointest Endosc 1995;41:421. 11. Sherman S, Lehman GA. ERCP and endoscopic induced pancreatitis. Pancreas 1991;6:350-67.
GASTROINTESTINAL ENDOSCOPY
71
Background: From a technological viewpoint, there have been few advances in retrograde pancreatography since the inception of ERCP despite remarkable advances in other radiographic techniques such as CT and magnetic resonance imaging. We describe a new method of pancreatographic imaging, endoscopic digital pancreatography, based on digital image intensification. Methods: Thirty-two nonconsecutive patients with various pancreatic diseases were entered into the study. Pancreatography was performed using standard technique with a non-ionic, low osmolarity contrast medium. In addition to conventional pancreatographic x-rays, digital images were acquired and then processed by means of computer-assisted technical modalities. Both pancreatograms were evaluated by two independent investigators. Results: Digital pancreatography was rated as superior to conventional pancreatography in terms of overall quality as well as specific parameters. Xray dosage was significantly higher and fluoroscopy time significantly longer for conventional pancreatography. Conclusion: Endoscopic digital pancreatography represents a significant advance, one with several important advantages compared with conventional pancreatography. Endoscopic retrograde pancreatography (ERP) has, since its introduction almost 30 years ago, had a major impact on the assessment of patients with pancreatic disorders.1-5 Despite the extensive application of ERP, both diagnostic and therapeutic, there have been relatively few novel changes in the procedure from a technologic point of view,4,5 espe-
Received August 30, 1996. For revision August 19, 1997. Accepted March 11, 1998. From the Pancreas Unit, Department of Gastroenterology and Radiology, Cardarelli Hospital, Naples, Italy. Reprint requests: G. Uomo, MD, Pancreas Unit, Department of Gastroenterology, Cardarelli Hospital, via Cardarelli 9, 80131 Napoli, Italy. Copyright © 1998 by the American Society for Gastrointestinal Endoscopy 0016-5107/98/$5.00 + 0 37/69/90175 VOLUME 48, NO. 1, 1998
cially by comparison with the evolution of other radiologic techniques for the study of the pancreas, such as CT and magnetic resonance. This report describes preliminary results with a new pancreatographic imaging method based on digital image intensification that which we call endoscopic digital pancreatography (EDP). X-ray exposure was also measured. In our study, digital pancreatograms were rated as superior in overall quality and a variety of specific parameters compared with conventional pancreatograms. PATIENTS AND METHODS Thirty-two nonconsecutive patients were entered into the study over an 18-month period (July 1994 to December 1995). Cannulation of the main duodenal papilla was achieved in 30 patients. There were 13 men and 17 women; mean age was 56 ± 15.6 years (range 24 to 77 years). The indications for visualization of the pancreatic duct (PD) were as follows: suspected chronic pancreatitis (16 patients), suspected pancreatic cancer (9 patients), idiopathic acute pancreatitis (3 patients), and pseudocyst post-acute pancreatitis (2 patients). Duodenoscopy was performed by one operator (G.U.) using standard technique and equipment.4 Non-ionic, low osmolarity contrast medium (iopamidol) was used in all cases. Digital acquisition of images was initiated when early PD injection was evident on fluoroscopy; total acquisition was usually completed in a matter of seconds and never required longer than thirty seconds. EDP was performed using a radioscopic/graphic unit (Siregraph C; Siemens, Erlangen, Germany) on line with a computer (Digitron 3VAI; Siemens Medical Service, Chicago, Ill.). Images were achieved in continuous mode with a rate of four frames per second (120 images in 30 seconds). The automatic exposure control features included fluoroscopically programmed radiography in which the correct parameters for each exposure were automatically calculated and set from the fluoroscopic parameters. The digital imaging chain provided an image (512 × 512 matrix size) with a 10-bit gray scale, series exposures, and automatic image optimization. Conventional ERP was also performed for each patient during the same session using the same radiographic unit. After EDP completion, images were processed by means of computer-assisted technical modalities including dynamic zoom, magnification, image-subtraction, automatic and manual adjustment of gray-scale, calibration parameters, and video-tape recording. Moreover, a special filter (Sobel, Digitron) was used when an enhancement of PD edges was considered to be helpful. All of these noted computer-assisted processing methods required just a few seconds: for example, filtration required 10 seconds. Each pancreatogram obtained with EDP and ERP was evaluated by two skilled independent observers who rated the overall quality of the images, as well as the view of the lumen and edges of the main pancreatic duct together with its secondary and tertiary branches. The images GASTROINTESTINAL ENDOSCOPY
67
G Uomo, R de Ritis, P Rabitti, et al.
Endoscopic digital pancreatography: An advance in pancreatic imaging?
Table 1. Numbers of patients with either poor-sufficient or excellent image ratings for overall quality, and lumen and edges of the main pancreatic duct, secondary and tertiary branches. Poor-sufficient
Excellent
5 15
25 (p = 0.006) 15
Overall quality EDP ERP Edge Poor-sufficient MPD EDP ERP Secondary branches EDP ERP Tertiary branches EDP ERP
Lumen Excellent
Poor-sufficient
Excellent
2 14
28 (p = 0.01) 16
2 10
28 (p = 0.01) 20
9 13
21 17
9 14
21 (p = 0.01) 16
9 12
21 (p = 0.01) 18
19 20
11 10
EDP, Endoscopic digital pancreatography; ERP, conventional retrograde pancreatography; MPD, main pancreatic duct.
were judged as poor-sufficient or excellent. When the evaluations of the two observers did not agree (four cases), a joint revaluation was carried out. X-ray exposure was measured in an additional six EDP and six ERP procedures not included in the study. Dosage was measured by means of termoluminescent dosimeters placed at the level of the thyroid and left eyelid of the endoscopist nearest to the x-ray table and of the patient. Fluoroscopy time was also recorded for both types of examination. Statistical analysis was performed using Student’s ttest, Fisher exact test, and Friedman test, as indicated. Significance was set at p < 0.05.
RESULTS Numbers of patients with either poor-sufficient or excellent ratings of EDP and ERP images with regard to overall quality and the lumen and edges of the main pancreatic duct together with its secondary and tertiary branch ducts are shown in Table 1. Examples of EDP images and corresponding standard ERP images are presented in Figures 1 through 3. EDP was especially useful in patients with pancreatic cancer. In seven patients with a confirmed diagnosis of malignancy, there were significant differences in the ratings for EDP versus ERP images in terms of overall quality, main pancreatic duct edge, and the edge and lumen of secondary ducts. However, differences in the several rating parameters did not reach statistical significance in the group of 13 patients with confirmed chronic pancreatitis. X-ray dosage was significantly higher for ERP compared with EDP at the level of eyelid both for patients (10.4 ± 2.7 mGy in ERP, 2.4 ± 0.9 mGy in EDP; p < 0.001) and operator (0.3 ± 0.2 mGy in ERP, 68
GASTROINTESTINAL ENDOSCOPY
0.07 ± 0.02 in EDP; p = 0.01). Fluoroscopy time was significantly longer for ERP (120 ± 21.5 seconds) than EDP (24 ± 4.2 seconds); p < 0.001). DISCUSSION Digital image intensifier radiology, a sophisticated technique that uses an x-ray image intensifier and a television system with a digitized video signal, is still undergoing relatively rapid development.6 It has been used in recent years for a broad range of examinations including barium-contrast studies of the gastrointestinal tract and several iodine-based contrast studies.6-8 Digital technique offers many advantages in comparison with conventional radiology.6-9 These include (1) an increase in image quality due to acquisition of a real-time series of images and various types of analysis of images (dynamic high-resolution fluoroscopic images, single-shot and series exposures on the monitor, images recorded on film); (2) post-processing image-handling capabilities (adjustment of gray-scale value and contrast, edge enhancement, magnification) that provide for optimal image presentation; (3) digital storage of images that is easier and less expensive than conventional methods of image archiving and also allows for image transfer in a communication network; and, (4) reduced radiation exposure because less radiation is required per exposure and there are fewer unsatisfactory exposures. Moreover, digital radiology requires far fewer standard x-ray films with a corresponding reduction in the environmental load associated with processing of chemical and silver halides. The major disadvantage of digital VOLUME 48, NO. 1, 1998
Endoscopic digital pancreatography: An advance in pancreatic imaging?
Figure 1. Cancer of the body of the pancreas confirmed at surgery. A, Conventional pancreatography demonstrates interruption and stenosis of the main pancreatic duct at the junction of the head and body (small arrows) and an irregular filling defect in the accessory duct (large arrow). B, EDP visualization in the same patient of the main pancreatic duct at level of body and tail also demonstrates irregular, longstenosis and better resolution of the secondary and tertiary branches in the head of the gland as well as better resolution of the stenosis in the accessory pancreatic duct. VOLUME 48, NO. 1, 1998
G Uomo, R de Ritis, P Rabitti, et al.
Figure 2. Cancer in the head of the gland that was confirmed at surgery. A, Conventional pancreatogram: the presence of contrast medium in the duodenum covers a suspected stenosis of the main pancreatic duct. B, EDP of same patient with subtraction of duodenal contents and use of an edge enhancement filter provides clear visualization of an irregular stenosis (arrows) of the main pancreatic duct.
GASTROINTESTINAL ENDOSCOPY
69
G Uomo, R de Ritis, P Rabitti, et al.
Endoscopic digital pancreatography: An advance in pancreatic imaging?
Figure 3. EDP with subtraction of osseous and gaseous interferences shows parenchymography of the tail of the gland. Surgical exploration revealed a pancreatic adenocarcinoma.
imaging relates to its limited spatial resolution which renders it less suitable for large input screens, such as in chest examination.6 The application of digital radiologic technique to pancreatography has been reported by Bickston et al.10 These investigators performed digital pancreatography in 10 patients and judged the images obtained to be superior to those achievable with conventional radiologic technique. In the present study, the availability of high-quality, real-time images with automatic optimization and post-processing modalities significantly improved the overall quality of pancreatographic images and displayed the main pancreatic duct together with its secondary and tertiary branches in greater detail. Through use of the dynamic acquisition capability of EDP we were able to monitor the progressive filling and opacification of the pancreatic ducts. This proved to be a crucial factor in the precise identification of communications between the ductal system and parenchymal cavities and also made it easier to recognize segmental ductal overfilling. Moreover, because of enhanced contrast and resolution, EDP recognized impending acinarization earlier than was possible by ERP (Fig. 3). Based on this last observation and the lower concentrations of contrast medium generally used in digital technique10 and considering that acinarization is an important risk factor in pancreatic inflammation associated with retrograde pancreatography,11 it is possible that EDP might reduce the frequency of procedure-related pancreatitis compared with ERP. 70
GASTROINTESTINAL ENDOSCOPY
In effect, EDP provides the endoscopist with more precise control of the injection phase of the procedure. The limited spatial resolution of EDP in comparison with ERP was evident during the acquisition phase of the procedure. However, a satisfactory result can be obtained by means of the post-processing modalities mentioned previously. Digital recording of images makes it possible to review the entire examination minute by minute as compared with the study of single x-ray pictures as is the case with conventional ERP. This can be especially advantageous when dealing with an evolving lesion such as a pseudocyst. From the economic standpoint, EDP costs around $200 per examination which is more or less in line with our cost for a traditional ERP. However, this is the operational cost and does not take into account the fact that a digital computerized radiology unit was already available in our hospital and does not include the costs for technical personnel and physicians fees. It is often necessary to provide referring physicians with “hard copy” images of pancreatographic images. With EDP, there is less need for duplicate images. By selecting images, it is possible to demonstrate the most significant findings with fewer images, which becomes a cost-saving factor. EDP significantly reduced x-ray exposure for both patients and operators when compared with ERP. In addition to a significant reduction in fluoroscopy time, there were fewer attempts to better demonstrate a finding by repeated injections. In addition to their high quality, the images recorded after filling is complete can be reviewed repeatedly without the need for further x-ray exposures. In conclusion, our experience and data suggest that EDP should be considered a true advance in pancreatic imaging, one that offers several important advantages in comparison with conventional pancreatography. Studies of larger series of patients are needed to fully investigate the potential advantages and limitations of this new technique including cost-effectiveness, potential to reduce procedurerelated complications, and enhancement of pancreatographic diagnosis. REFERENCES 1. Axon ATR, Classen M, Cotton PB. Pancreatography in chronic pancreatitis: international definitions. Gut 1984;25:1107-12. 2. Geenen JE, ASGE distinguished lecture. Endoscopic therapy of pancreatic disease: a new horizon. Gastrointest Endosc 1988;34:386-91. 3. Schapiro RH. ERCP in the diagnosis of pancreatic and biliary disease. In: Jacobson IM, eds. ERCP, diagnostic and therapeutic applications. Elseiver, New York, 1989;9-39. 4. Siegel JH, editor. Endoscopic retrograde cholangiopancreVOLUME 48, NO. 1, 1998
atography. Technique, diagnosis and therapy. New York: Raven Press; 1992. 5. Huibregtse K, Smits ME. Endoscopic management of disease of the pancreas. Am J Gastroenterol 1994;89:S67-77. 6. Lehmann K, Busch HP, Georgi M. Digital fluoroscopic radiology: evaluation of clinical impact. Eur J Radiol 1993;17:3-7. 7. Gingold EL, Tucker DM, Barnes GT. Computed radiography: user-programmable features and capabilities. J Digital Imag 1994;7:113-8.
VOLUME 48, NO. 1, 1998
8. Busch HP, Georgi M. Digital radiographs. Quality assurance and radiation protection. Radiol Diagn 1992;33:321-4. 9. Torrent M, Moreau M, Deschamps D, Capron J. Digital imaging in conventional radiology. Medica Mundi 1995;40:3-11. 10. Bickston SJ, Balfe DM, Brink JA, Aliperti G. Digital pancreatography is superior to conventional pancreatography [abstract]. Gastrointest Endosc 1995;41:421. 11. Sherman S, Lehman GA. ERCP and endoscopic induced pancreatitis. Pancreas 1991;6:350-67.
GASTROINTESTINAL ENDOSCOPY
71