Ultrasound
Visualization of the Bile Ducts Using Focused Ultrasound 1 Michael H. Reid, Ph.D., M.D. The feasibility of using focused ultrasound for studying structures as small as biliary ducts and correlative radiological and ultrasonic studies of the common and larger hepatic ducts are presented. Patients whose bile ducts were visualized on intravenous or T-tube cholangiography or Oral cholecystography underwent radiography witn a metallic skin marker along the right anterior lower costal margin.' The marker was then correlated with ultrasonic observations of multiple oblique scans parallel to the lower costal margin. The common duct could be identified in 15 of 21 patients. Ultrasonic visualization circumvents the problems of elevated serum bilirubin, iodine contrast media sensitivity, and pregnan-
cy. Bile ducts, ultrasound
INDEX TERMS:
Radiology 118: 155-158, January 1976
biliary system without iodinated contrast media is useful in patients with serum bilirubin over 4 mg/ 100 ml or with known iodine contrast media sensitivity. Occasionally, the risks of x irradiation may not be justified, e.g., during pregnancy. An ultrasonic approach to bile duct Visualization circumvents these problems and this paper discusses the practicality of using ultrasound for studying small structures. Correlative radiological and ultrasonic studies of the common and larger hepatic ducts are presented.
D
ULTRASONIC RESOLUTION
ELINEATION OF THE
B-scan resolution is limited by mechanical factors (scanning arm position and tracking), echo timing (depth resolution), and transducer diameter (near-field sound beam width). Depth resolution is directly proportional to
W=L22~P/D
w
~."~~-----_~_J
-------
~---------r----------
~p--1
n
em
MHz »2/4 em
f
1.3 2.0 5.6
2.54
2.54
3.5
3.5
38
38
F
em
10
10
5
•
em
0.7
0.2
0.1
Fig. 1. Transducer-lens focusing parameters. transducer diameter; f transducer freD ultrasound wavelength in tissue at quency; A frequency f; W = half-power beam width at the focal point.
=
=
=
2 Fig. 3. Transducer-iens measured focusing characteristics, in line palrs/crn resolved at 10-db dynamic range vs. range in centimeters. A. Nonfocused, 2.0-MHz, 1.3-cm diameter transducer. B-D. 3.5-MHz, 2.54-cm diameter transducer with lenses of focal length 5,7, and 10 cm, respectively.
1 From the Department of Radiology, University of California School of Medicine, San Francisco, Calif. Presented at the Sixtieth Scientific Assembly and Annual Meeting of the Radiological Society of North America, Chicago, III., Dec. 1-6, 1974. elk
155
Fig. 4. Patient L.D. Oral cholecystogram, showing an opacified gallbladder and a skin marker at the lower costal marqin. Fig. 5. Patient L.D. Linear tomogram shows an opacified gallbladder and faint visualization of the common bile duct. Fig. 6. Patient L.D. Ultrasound scan along the right lower costal margin with a lens of focal length 6 cm. GB = gallbladder; BD = bile duct; IVC = inferior vena cava; A = aorta; PV = portal vein. All scale markings arein centimeters. Fig. 7. Patient L.D. Ultrasound scan along the right lower costal margin with a lens of focal length 6 cm and the transducer angled 100 superiorly.
Fig. 8. Patient R.W. T-tube cholangiogram, showing the lower costal margin skin marker. Fig: 9. Patient R.W. Ultrasound scan along the right lower costal margin with a lens of focal length 7 cm and the transducer angled 100 superiorly.
Fig. 10. superiorly.
Patient R.W. Ultrasound scan along the right lower costal margin with a lens of focal length 8 cm and the transducer angled 150
156
157
VISUALIZATION OF THE BILE DUCTS
Vol. 118
Ultrasound
the precision of echo timing and is a function of the attenuation and velocity-dispersion characteristics of the tissue and the electronic pulse-shaping and amplification of detected echoes. Practically, for a given ultrasonic frequency, the depth resolution is on the order of two or three wavelengths. Mechanical resolution is difficult to evaluate because it depends on the length and excursions of the scanning arm and the precision in sensing the transducer position and angle. For small sector scans of points less than 8 or 10 cm from the transducer, mechanical resolution is usually better than 2mm. Ultrasonic near-field beam width, which is approximately the diameter of the transducer, is usually the limiting factor in resolution. In compound B-scanning it limits resolution in all directions because structures are purposely scanned from disparate positions; this is in contrast to single-pass scans where, at least, depth resolution is better preserved. Beam width may be reduced by employing small-diameter transducers or by focusing the beam. The minimization of transducer diameter is limited by decreased sensitivity to returning echoes, decreased length of the collimated ultrasonic beam (near-field), and finally, diffraction phenomena as the transducer diameter approaches the sound wavelength. EXPERIMENTAL DESIGN
To be compatible with available B-scan equipment and to allow selection of various depths of focus, a flat 3.5-MHz transducer 2.54 cm in diameter with a set of snap-on Lucite plastic lenses was employed in this study.? This choice of diameter represented a compromise of aperture size for good focusing, depth of focus, sensitivity, and availability of industrial transducers. The 3.5-MHz frequency has a wavelength of 0.4 mm in tissue. This gives a predicted depth resolution and halfpower focal spot diameter of 1 to 2 mm when allowing for approximately 1 cm of depth of focus at focal points ranging from 5 to 12 em. Tissue inhomogeneity degrades the focus by an amount which is, as yet, unknown. The ratio of ultrasonic intensity at the focal point to that at the skin varies from 0.0008 % for a 10-cm focal length to 10% for a 5-cm focal length, assuming an average tissue attenuation of 2 db/cm-MHz. Figure 1 illustrates some of the above points. Figure 2 shows the transducer and two of the lenses of the set, covering 5-12-cm focal lengths. Figure 3 is a set of curves representing resolution (line pairs/ern) VS. transducer-object distance for a standard nonfocused 2.0MHz transducer 1.3 cm in diameter (Fig. 3, A) and for the transducer-lens combination of this study (Fig. 3, B-D). These plots were obtained for a 10-db dynamic range, i.e., the echo amplitude when focused over a line 2 Manufactured by Aerotec Corp., Lewiston, Pa. Lenses were machined from Lucite plastic which is readily available and acoustically appropriate for interfacing the transducer to tissue.
Fig. 11. Patient P.C. Intravenous cholangiogram, showing the lower costal margin skin marker. Fig. 12. Patient P.C. Ultrasound scan along the right lower costal margin with lens of focal length 6 cm and the transducer angled 10° superiorly.
was 10 db larger than that obtained over a space for the particular line spacing used. CLINICAL DATA
Patients whose bile ducts were visualized on intravenous or T-tube cholangiography or oral cholecystography underwent radiography with a metallic marker along the right anterior lower costal margin. This skin marker, coupled with the tomographic level of duct visualization,
158
MICHAEL H. REID
was then correlated with the ultrasonic observations of multiple oblique scans parallel to the costal margin. This scanning technique had three advantages: (a) it generally cut perpendicularly across the porta hepatis, so that the common duct appeared as a sonolucent circle; (b) it avoided rib-crossing artifacts while still scanning much of the liver in suspended respiration (inspiration); and (c) the porta hepatis was maintained approximately at the focal point of the lens for this scan configuration. An initial scan with a nonfocused tranducer identified the region of the porta hepatis. The appropriate lens was then selected and more precise scans were made of the area. The common duct and portions of the larger hepatic and cystic ducts could be identified in 15 of 21 patients. Identification of intrahepatic ducts was successful only in a minority of cases (4 patients) and is related to the problems of compound scanning and of display without gray scale. Figures 4-12 illustrate correlative radiographic and ultrasonic visualization of the bile ducts. The radiographs in Figures 4, 5, 8, and 11 are presented with
January 1976
right-left reversal to simplify correlation with the corresponding ultrasound scans. CONCLUSIONS
The results of this brief investigation and those of Taylor et al.3 demonstrate the feasibility of employing focused ultrasound for visualizing small structures such as the biliary ducts. Further improvements in signal processing and gray-scale displays will certainly provide accurate and consistant identification of the biliary system. Department of Radiology University of California School of Medicine San Francisco, Calif. 94143
3 After submission of the abstract for this paper, Taylor et a/. reported much the same results in the Journal of Clinical Ultrasound (Vol. 2, Jun 1974, pp. 105-116). Credit must be ~liven to Taylor's sophisticated gray scale and the clarity of his scans, which exceed those possible with the equipment currently available to me.
Visualization of the Bile Ducts Using Focused Ultrasound 1 Michael H. Reid, Ph.D., M.D. The feasibility of using focused ultrasound for studying structures as small as biliary ducts and correlative radiological and ultrasonic studies of the common and larger hepatic ducts are presented. Patients whose bile ducts were visualized on intravenous or T-tube cholangiography or Oral cholecystography underwent radiography witn a metallic skin marker along the right anterior lower costal margin.' The marker was then correlated with ultrasonic observations of multiple oblique scans parallel to the lower costal margin. The common duct could be identified in 15 of 21 patients. Ultrasonic visualization circumvents the problems of elevated serum bilirubin, iodine contrast media sensitivity, and pregnan-
cy. Bile ducts, ultrasound
INDEX TERMS:
Radiology 118: 155-158, January 1976
biliary system without iodinated contrast media is useful in patients with serum bilirubin over 4 mg/ 100 ml or with known iodine contrast media sensitivity. Occasionally, the risks of x irradiation may not be justified, e.g., during pregnancy. An ultrasonic approach to bile duct Visualization circumvents these problems and this paper discusses the practicality of using ultrasound for studying small structures. Correlative radiological and ultrasonic studies of the common and larger hepatic ducts are presented.
D
ULTRASONIC RESOLUTION
ELINEATION OF THE
B-scan resolution is limited by mechanical factors (scanning arm position and tracking), echo timing (depth resolution), and transducer diameter (near-field sound beam width). Depth resolution is directly proportional to
W=L22~P/D
w
~."~~-----_~_J
-------
~---------r----------
~p--1
n
em
MHz »2/4 em
f
1.3 2.0 5.6
2.54
2.54
3.5
3.5
38
38
F
em
10
10
5
•
em
0.7
0.2
0.1
Fig. 1. Transducer-lens focusing parameters. transducer diameter; f transducer freD ultrasound wavelength in tissue at quency; A frequency f; W = half-power beam width at the focal point.
=
=
=
2 Fig. 3. Transducer-iens measured focusing characteristics, in line palrs/crn resolved at 10-db dynamic range vs. range in centimeters. A. Nonfocused, 2.0-MHz, 1.3-cm diameter transducer. B-D. 3.5-MHz, 2.54-cm diameter transducer with lenses of focal length 5,7, and 10 cm, respectively.
1 From the Department of Radiology, University of California School of Medicine, San Francisco, Calif. Presented at the Sixtieth Scientific Assembly and Annual Meeting of the Radiological Society of North America, Chicago, III., Dec. 1-6, 1974. elk
155
Fig. 4. Patient L.D. Oral cholecystogram, showing an opacified gallbladder and a skin marker at the lower costal marqin. Fig. 5. Patient L.D. Linear tomogram shows an opacified gallbladder and faint visualization of the common bile duct. Fig. 6. Patient L.D. Ultrasound scan along the right lower costal margin with a lens of focal length 6 cm. GB = gallbladder; BD = bile duct; IVC = inferior vena cava; A = aorta; PV = portal vein. All scale markings arein centimeters. Fig. 7. Patient L.D. Ultrasound scan along the right lower costal margin with a lens of focal length 6 cm and the transducer angled 100 superiorly.
Fig. 8. Patient R.W. T-tube cholangiogram, showing the lower costal margin skin marker. Fig: 9. Patient R.W. Ultrasound scan along the right lower costal margin with a lens of focal length 7 cm and the transducer angled 100 superiorly.
Fig. 10. superiorly.
Patient R.W. Ultrasound scan along the right lower costal margin with a lens of focal length 8 cm and the transducer angled 150
156
157
VISUALIZATION OF THE BILE DUCTS
Vol. 118
Ultrasound
the precision of echo timing and is a function of the attenuation and velocity-dispersion characteristics of the tissue and the electronic pulse-shaping and amplification of detected echoes. Practically, for a given ultrasonic frequency, the depth resolution is on the order of two or three wavelengths. Mechanical resolution is difficult to evaluate because it depends on the length and excursions of the scanning arm and the precision in sensing the transducer position and angle. For small sector scans of points less than 8 or 10 cm from the transducer, mechanical resolution is usually better than 2mm. Ultrasonic near-field beam width, which is approximately the diameter of the transducer, is usually the limiting factor in resolution. In compound B-scanning it limits resolution in all directions because structures are purposely scanned from disparate positions; this is in contrast to single-pass scans where, at least, depth resolution is better preserved. Beam width may be reduced by employing small-diameter transducers or by focusing the beam. The minimization of transducer diameter is limited by decreased sensitivity to returning echoes, decreased length of the collimated ultrasonic beam (near-field), and finally, diffraction phenomena as the transducer diameter approaches the sound wavelength. EXPERIMENTAL DESIGN
To be compatible with available B-scan equipment and to allow selection of various depths of focus, a flat 3.5-MHz transducer 2.54 cm in diameter with a set of snap-on Lucite plastic lenses was employed in this study.? This choice of diameter represented a compromise of aperture size for good focusing, depth of focus, sensitivity, and availability of industrial transducers. The 3.5-MHz frequency has a wavelength of 0.4 mm in tissue. This gives a predicted depth resolution and halfpower focal spot diameter of 1 to 2 mm when allowing for approximately 1 cm of depth of focus at focal points ranging from 5 to 12 em. Tissue inhomogeneity degrades the focus by an amount which is, as yet, unknown. The ratio of ultrasonic intensity at the focal point to that at the skin varies from 0.0008 % for a 10-cm focal length to 10% for a 5-cm focal length, assuming an average tissue attenuation of 2 db/cm-MHz. Figure 1 illustrates some of the above points. Figure 2 shows the transducer and two of the lenses of the set, covering 5-12-cm focal lengths. Figure 3 is a set of curves representing resolution (line pairs/ern) VS. transducer-object distance for a standard nonfocused 2.0MHz transducer 1.3 cm in diameter (Fig. 3, A) and for the transducer-lens combination of this study (Fig. 3, B-D). These plots were obtained for a 10-db dynamic range, i.e., the echo amplitude when focused over a line 2 Manufactured by Aerotec Corp., Lewiston, Pa. Lenses were machined from Lucite plastic which is readily available and acoustically appropriate for interfacing the transducer to tissue.
Fig. 11. Patient P.C. Intravenous cholangiogram, showing the lower costal margin skin marker. Fig. 12. Patient P.C. Ultrasound scan along the right lower costal margin with lens of focal length 6 cm and the transducer angled 10° superiorly.
was 10 db larger than that obtained over a space for the particular line spacing used. CLINICAL DATA
Patients whose bile ducts were visualized on intravenous or T-tube cholangiography or oral cholecystography underwent radiography with a metallic marker along the right anterior lower costal margin. This skin marker, coupled with the tomographic level of duct visualization,
158
MICHAEL H. REID
was then correlated with the ultrasonic observations of multiple oblique scans parallel to the costal margin. This scanning technique had three advantages: (a) it generally cut perpendicularly across the porta hepatis, so that the common duct appeared as a sonolucent circle; (b) it avoided rib-crossing artifacts while still scanning much of the liver in suspended respiration (inspiration); and (c) the porta hepatis was maintained approximately at the focal point of the lens for this scan configuration. An initial scan with a nonfocused tranducer identified the region of the porta hepatis. The appropriate lens was then selected and more precise scans were made of the area. The common duct and portions of the larger hepatic and cystic ducts could be identified in 15 of 21 patients. Identification of intrahepatic ducts was successful only in a minority of cases (4 patients) and is related to the problems of compound scanning and of display without gray scale. Figures 4-12 illustrate correlative radiographic and ultrasonic visualization of the bile ducts. The radiographs in Figures 4, 5, 8, and 11 are presented with
January 1976
right-left reversal to simplify correlation with the corresponding ultrasound scans. CONCLUSIONS
The results of this brief investigation and those of Taylor et al.3 demonstrate the feasibility of employing focused ultrasound for visualizing small structures such as the biliary ducts. Further improvements in signal processing and gray-scale displays will certainly provide accurate and consistant identification of the biliary system. Department of Radiology University of California School of Medicine San Francisco, Calif. 94143
3 After submission of the abstract for this paper, Taylor et a/. reported much the same results in the Journal of Clinical Ultrasound (Vol. 2, Jun 1974, pp. 105-116). Credit must be ~liven to Taylor's sophisticated gray scale and the clarity of his scans, which exceed those possible with the equipment currently available to me.
