- Email: [email protected]
A simple device for respiratory gating for the MRI of laboratory animals
Magnelic Resonance Imaging, Vol. I I. pp. 897-901, Printed in the USA. All rights reserved.
1993 Copyright 0
0730-725X/93 $6.00 + .OO 1993 Pergamon Press Ltd.
l Technical Note
A SIMPLE DEVICE FOR RESPIRATORY GATING FOR THE MRI OF LABORATORY ANIMALS NEWMAN
G. BURDETT,
T. ADRIAN
CARPENTER,
AND LAURENCE
D. HALL
Herchel Smith Laboratory for Medicinal Chemistry, University of Cambridge School of Clinical Medicine, University Forvie Site, Cambridge, UK CB2 2PZ Respiratory motion must be overcome if MRI of the abdomen, even at the lowest resolution, is to be performed satisfactorily. A simple and reliable respiratory gating device, based on the interruption of an infrared (IR) optical beam is described. This device has the advantage that gating is based on the position of the chest as opposed to its velocity, and that it can be used without degrading the radiofrequency isolation of a Faraday cage. Its use in animal MRI is illustrated by high resolution (200 am) images of in vivo rat liver and kidney. Keywords: MRI; Respiratory gating; Kidney; Liver; Gut motion.
tively, specialised motion insensitive pulse sequences can be used, such as Respiratory Ordered Phase Encoding (ROPE);6 these remove most phase-encode artifacts without increasing the scan time, but do not reduce image blurring, nor the variability of signal intensity. To remove both the latter, some method of synchronising data acquisition with respiration must be used, such as a respiratory gatingjtriggering. This has the drawback of making the repetition time (TR) dependent on some aspect of the respiratory cycle, so that the actual TR will be the chosen delay plus some varying length of time until the respiratory cycle returns to the point that causes triggering. This causes a lengthening of the total acquisition time. Careful choice of the basic TR to be just less than some integral multiple of the respiratory cycle ensures efficient triggering, but even so, the average acquisition time is generally doubled. If imaging time needs to be kept to a minimum, it is generally better to reduce the number of averages rather than to remove respiratory gating.’ Respiratory gating also causes problems in pulse sequences where TR is important, such as saturation recovery, although this can be avoided by use of alternative sequences such as inversion recovery. In order to use a defined point of the respiratory cycle as a synchronisation trigger for MRI, some means of measuring it must be utilised. Several methods for implementing respiratory gating already exist, the most common being pressure transducers, which record the
INTRODUCTION
In preparation for a major physiological study involving magnetic resonance imaging (MRI) of the rat kidney, we required a protocol and equipment for routine imaging. Few studies on laboratory rats’,* have been reported, due to problems with increased respiratory rate and decreased organ size. As each set of MR images takes on the order of minutes to acquire, any motion of the subject during that time leads to a blurring of the image. This blurring is compounded when twodimensional Fourier reconstruction is used, because any regular movement, such as respiration, which, regardless of the direction of actual motion, gives rise to aliased “ghost” images in the phase-encode direction, obscures anatomical detail and makes the image difficult to interpret.3 Even in the absence of obvious motion artifacts, respiratory motion can alter signal intensity, making measurements of Ti- and T2-relaxation times and diffusion coefficients inaccurate,4 in extreme cases giving rise to apparently negative values.5 Placing the rat supine minimises these effects by minimising organ motion; however,this is insufficient for the production of good quality images. Several methods have been used previously to overcome artifacts due to motion. The simplest of these is to ask the patient to refrain from breathing during the scan. This is impractical for all but the shortest scans (5 lmin) and is unfeasible for animal studies. AlternaRECEIVED 11/27/92;
Address
ACCEPTED 3/3/93. 897
correspondence
to Newman
G. Burdett.
898
Magnetic Resonance Imaging 0 Volume I 1, Number 6, 1993
changes in pressure produced in a bag of fluid by movement of the chest during inspiration.234 These devices are, however, often magnetic and expensive. Motion of a loop of wire in a magnetic field causes a current to be induced, which is proportional to the rate of change of the magnetic flux in the loop; by placing the loop on the chest of the subject the motion of the chest can be monitored.8 This method has the disadvantage of not relating triggering signal to chest position, but, rather, to the rate of motion of the chest, which may not correspond exactly. Furthermore, it also places the wires in close proximity tot he area of interest which can give rise to susceptibility artifacts. A method of respiratory monitoring commonly used in small animal surgery relies on detecting changes in temperature caused by inhalation and exhalation; this has been tried in conjunction with MRI but was found to be unsatisfactory.’ A more elegant approach is to use some intrinsic part of the detected MR signal to monitor the breathing cycle,‘OS1’and a few methods utilising this have been demonstrated, one using transverse profiles to record chest position and the other changes in the loading of the coil throughout the cycle. In this study, a device has been developed which is based on an infrared optico-interrupt switch; it appears to be an inexpensive and simple alternative to the above methods; furthermore, it can be used in conjunction with an MRI system located inside the Faraday cage since no electrical wires pass through the cage causing radiofrequency (RF) noise leakage. In order to assess the effectiveness of the gating and the spatial resolution that could be achieved using it, normal anesthetized rats
were imaged, with and without gating and the image quality compared. In the course of these studies, it has also been shown that intraperitoneal (IP) injection of a cocktail of Hypnorm (Janssen Pharmaceutical Ltd., Oxford, UK) and Hypnovel (Roche, Welwyn Garden City, UK) abolishes gut motion, thereby removing a secondary source of motion artifacts. Previously, intravenous injection of glucagon has been used to achieve this reduction in peristaltic motion.12 MATERIALS
AND METHODS
The gating trigger (see Fig. 1) consists of a nylon plunger, located in a Plexiglas former, which rests on the animal’s chest and during inspiration, interrupts an IR beam carried by fibre optic cables. The optical fibres pass outside the faraday cage, where the IR beam is detected and used to provide a triggering pulse for each excitation of the image acquisition (the IR emitter, detector, and fibre optic cables were supplied by Maplin Electronics plc.). Male Wistar (200-450 g) rats were anesthetized by IP injection of 0.3 ml/lOOg bodyweight Hypnorm/Hypnovel cocktail (diluted in the ratio 1: 1:2 with sterilised water). Images were acquired in a 3 I cm, horizontal bore 2.4 T superconducting magnet (Oxford Instruments Ltd.) using a Bruker Medezintechnik Biospec 11console and 20 cm internal diameter gradient coils built in house. The rats were placed in a 5 cm diameter half birdcage resonator13 tuned to 100 MHz. Target images were made using sagittal and transverse pilot scans. The images were acquired using a
ating Box
Rat’s Abdomen
Fig. 1. Schematic diagram of the respiratory plunger up, interrupting the IR beam.
gating trigger used. Inspiration
causes the rat’s abdomen
to rise, forcing the nylon
Simple device for respiratory gating 0 N.G. BURDETTET AL.
899
-I
Fig. i. (Coronal slices through the abdomen of a male Wistar rat, ventrally to dorsally, with (a)-(d), and without (e)-(h) I,espiratory gati ng. mages (a)-(d) show clearer anatomical detail with less “aliasing” than images (e)-(h). (Figure continues)
spin-echo sequence with a 7 cm field of view, 2 mm slice thickness, 2 signal averages, a TR of either 1500 msec
(ungated) or 700 msec (+ a gating delay) to make the actual TR of the gated images approximately equal, a TE of 20 msec in all, and a 256 x 256 image matrix. RESULTS Figure 2 shows a set of coronal images acquired with (a)-(d), and without, (e)-(h), respiratory gating. The liver and kidney can be clearly seen in (c) and (d), along with the stomach and diaphragm. The aorta, vena cava, and vessels branching of these can also be seen in these images. Images (e)-(h) correspond spatially to images (a)-(d), but as they were acquired without gating the anatomical detail of images (a)-(d) is lacking; images (e) and (f) show quite good anatomi-
cal detail as they correspond to slices through positions just above the spine and through the spine, where there is little or no respiratory motion. As the image plane moves up through the body of the rat, respiratory motion increases, exacerbating motion artifacts and degrading image quality, as can be seen in the ungated images. CONCLUSIONS In summary, we present a technique for obtaining good quality MR images of the rat abdomen in vivo. The technique reduces artifacts due to motion sufficiently to allow MRI studies to be performed, thereby providing noninvasive access to animal models of the pathology and physiology of these organs in vivo. Its application to human subjects should be easier than in
Magnetic
900
Resonance
Imaging
Fig.
small animals due to the longer respiratory the larger range of movement.
cycle and
Acknowledgmenfs- It is a pleasure to thank Dr. Herchel Smith for a generous endowment (TAC, LDH) and for a research studentship (NGB), and Clifford Bunch for his technical assistance.
REFERENCES Reeker, F.; Rubben, H.; Bex, A.; Constantinides, C. Morphological changes following ESWL in the rat kidney. Ural. Res. 111229-233; 1989. Brauer, M.; Towner, R.A.; Renaud, 1.; Janzen, E.G.; Foxall, D.L. In vivo proton nuclear magnetic resonance imaging and spectroscopic studies of halocarbon-induced liver damage. Magn. Reson. Med. 9:229-239; 1989. Axel, L.; Summers, R.M.; Kressel, H.Y.; Charles, C. Respiratory effects in two-dimensional Fourier transform MR imaging. Radiology 160:795-801; 1986.
0 Volume
11, Number 6, 1993
2 continued.
4. Thompsen,
C.; Henriksen, 0.; Ring, P. In vivo measurements of relaxation processes in the human liver by MRI. The role of respiratory gating/triggering. Magn. Reson. Imaging 6:431-436; 1988. 5. Aisen, A.M.; Glazer, G.M.; Carson, P.L.; Hearshen, D.O. Motion artifacts in quantitative magnetic resonance imaging. Magn. Reson. Imaging 4:207-213; 1986. 6. Bailes, D.; Gilderdale, D.; Bydder, G.M.; Collins, A.G.; Firmin, D. ROPE a method for reducing respiratory motion artifacts in imaging. J. Comput. Assist. Tomogr. 9: 835-842; 1985. Erlichman, M. Surface/specialty coil devices and gating techniques in magnetic resonance imaging. Health Techno/. Assess. Rep. 3:1-23; 1990. Van Bruggen, N.; Syha, J.; Busza, A.L.; King, M.D.; Stamp, G.W.; Williams, S.R.; Gadian, D.G. Identification of tumor hemorrhage in an animal model using spin echoes and gradient echoes. Magn. Reson. Med. 15:121127; 1990. Ehman, R.L.; McNamara, M.T.; Pallack, M.; Hricak,
Simple device for respiratory gating 0 N.G. BURDETT ET AL. H.; Higgins, C.B. Magnetic resonance imaging with respiratory gating. Techniques and advantages. Am. J. Roentgenol. 143: 1175-l 182; 1984. 10. Cuppen, J.J.M.; Groen, J.; In Den Kleef, J.; Tuithof, H.A. Reduction of artifacts by data post processing. In: Book of abstracts: Fourth Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA: SMRM; 1985:962-963. 11. Buikman, D.; Helzel, T.; Roschcmann, P. The RF coil
901
as a sensitive motion detector. Magn. Reson. Imaging 6:281-289;
1988. 12. Shaff, M.I.; Tarr, R.W.; Partain,
CL.; James, A.E. Computed tomography and magnetic resonance imaging of the acute abdomen. Surg. Clin. North Am. 68: 233-254; 1988. 13. Ballon, D.; Graham, M.C.; Miodownik, S.; Koutcher, J.A. A 64 MHz half-birdcage resonator for clinical imaging. J. Magn. Reson. 90:131-140; 1990.
1993 Copyright 0
0730-725X/93 $6.00 + .OO 1993 Pergamon Press Ltd.
l Technical Note
A SIMPLE DEVICE FOR RESPIRATORY GATING FOR THE MRI OF LABORATORY ANIMALS NEWMAN
G. BURDETT,
T. ADRIAN
CARPENTER,
AND LAURENCE
D. HALL
Herchel Smith Laboratory for Medicinal Chemistry, University of Cambridge School of Clinical Medicine, University Forvie Site, Cambridge, UK CB2 2PZ Respiratory motion must be overcome if MRI of the abdomen, even at the lowest resolution, is to be performed satisfactorily. A simple and reliable respiratory gating device, based on the interruption of an infrared (IR) optical beam is described. This device has the advantage that gating is based on the position of the chest as opposed to its velocity, and that it can be used without degrading the radiofrequency isolation of a Faraday cage. Its use in animal MRI is illustrated by high resolution (200 am) images of in vivo rat liver and kidney. Keywords: MRI; Respiratory gating; Kidney; Liver; Gut motion.
tively, specialised motion insensitive pulse sequences can be used, such as Respiratory Ordered Phase Encoding (ROPE);6 these remove most phase-encode artifacts without increasing the scan time, but do not reduce image blurring, nor the variability of signal intensity. To remove both the latter, some method of synchronising data acquisition with respiration must be used, such as a respiratory gatingjtriggering. This has the drawback of making the repetition time (TR) dependent on some aspect of the respiratory cycle, so that the actual TR will be the chosen delay plus some varying length of time until the respiratory cycle returns to the point that causes triggering. This causes a lengthening of the total acquisition time. Careful choice of the basic TR to be just less than some integral multiple of the respiratory cycle ensures efficient triggering, but even so, the average acquisition time is generally doubled. If imaging time needs to be kept to a minimum, it is generally better to reduce the number of averages rather than to remove respiratory gating.’ Respiratory gating also causes problems in pulse sequences where TR is important, such as saturation recovery, although this can be avoided by use of alternative sequences such as inversion recovery. In order to use a defined point of the respiratory cycle as a synchronisation trigger for MRI, some means of measuring it must be utilised. Several methods for implementing respiratory gating already exist, the most common being pressure transducers, which record the
INTRODUCTION
In preparation for a major physiological study involving magnetic resonance imaging (MRI) of the rat kidney, we required a protocol and equipment for routine imaging. Few studies on laboratory rats’,* have been reported, due to problems with increased respiratory rate and decreased organ size. As each set of MR images takes on the order of minutes to acquire, any motion of the subject during that time leads to a blurring of the image. This blurring is compounded when twodimensional Fourier reconstruction is used, because any regular movement, such as respiration, which, regardless of the direction of actual motion, gives rise to aliased “ghost” images in the phase-encode direction, obscures anatomical detail and makes the image difficult to interpret.3 Even in the absence of obvious motion artifacts, respiratory motion can alter signal intensity, making measurements of Ti- and T2-relaxation times and diffusion coefficients inaccurate,4 in extreme cases giving rise to apparently negative values.5 Placing the rat supine minimises these effects by minimising organ motion; however,this is insufficient for the production of good quality images. Several methods have been used previously to overcome artifacts due to motion. The simplest of these is to ask the patient to refrain from breathing during the scan. This is impractical for all but the shortest scans (5 lmin) and is unfeasible for animal studies. AlternaRECEIVED 11/27/92;
Address
ACCEPTED 3/3/93. 897
correspondence
to Newman
G. Burdett.
898
Magnetic Resonance Imaging 0 Volume I 1, Number 6, 1993
changes in pressure produced in a bag of fluid by movement of the chest during inspiration.234 These devices are, however, often magnetic and expensive. Motion of a loop of wire in a magnetic field causes a current to be induced, which is proportional to the rate of change of the magnetic flux in the loop; by placing the loop on the chest of the subject the motion of the chest can be monitored.8 This method has the disadvantage of not relating triggering signal to chest position, but, rather, to the rate of motion of the chest, which may not correspond exactly. Furthermore, it also places the wires in close proximity tot he area of interest which can give rise to susceptibility artifacts. A method of respiratory monitoring commonly used in small animal surgery relies on detecting changes in temperature caused by inhalation and exhalation; this has been tried in conjunction with MRI but was found to be unsatisfactory.’ A more elegant approach is to use some intrinsic part of the detected MR signal to monitor the breathing cycle,‘OS1’and a few methods utilising this have been demonstrated, one using transverse profiles to record chest position and the other changes in the loading of the coil throughout the cycle. In this study, a device has been developed which is based on an infrared optico-interrupt switch; it appears to be an inexpensive and simple alternative to the above methods; furthermore, it can be used in conjunction with an MRI system located inside the Faraday cage since no electrical wires pass through the cage causing radiofrequency (RF) noise leakage. In order to assess the effectiveness of the gating and the spatial resolution that could be achieved using it, normal anesthetized rats
were imaged, with and without gating and the image quality compared. In the course of these studies, it has also been shown that intraperitoneal (IP) injection of a cocktail of Hypnorm (Janssen Pharmaceutical Ltd., Oxford, UK) and Hypnovel (Roche, Welwyn Garden City, UK) abolishes gut motion, thereby removing a secondary source of motion artifacts. Previously, intravenous injection of glucagon has been used to achieve this reduction in peristaltic motion.12 MATERIALS
AND METHODS
The gating trigger (see Fig. 1) consists of a nylon plunger, located in a Plexiglas former, which rests on the animal’s chest and during inspiration, interrupts an IR beam carried by fibre optic cables. The optical fibres pass outside the faraday cage, where the IR beam is detected and used to provide a triggering pulse for each excitation of the image acquisition (the IR emitter, detector, and fibre optic cables were supplied by Maplin Electronics plc.). Male Wistar (200-450 g) rats were anesthetized by IP injection of 0.3 ml/lOOg bodyweight Hypnorm/Hypnovel cocktail (diluted in the ratio 1: 1:2 with sterilised water). Images were acquired in a 3 I cm, horizontal bore 2.4 T superconducting magnet (Oxford Instruments Ltd.) using a Bruker Medezintechnik Biospec 11console and 20 cm internal diameter gradient coils built in house. The rats were placed in a 5 cm diameter half birdcage resonator13 tuned to 100 MHz. Target images were made using sagittal and transverse pilot scans. The images were acquired using a
ating Box
Rat’s Abdomen
Fig. 1. Schematic diagram of the respiratory plunger up, interrupting the IR beam.
gating trigger used. Inspiration
causes the rat’s abdomen
to rise, forcing the nylon
Simple device for respiratory gating 0 N.G. BURDETTET AL.
899
-I
Fig. i. (Coronal slices through the abdomen of a male Wistar rat, ventrally to dorsally, with (a)-(d), and without (e)-(h) I,espiratory gati ng. mages (a)-(d) show clearer anatomical detail with less “aliasing” than images (e)-(h). (Figure continues)
spin-echo sequence with a 7 cm field of view, 2 mm slice thickness, 2 signal averages, a TR of either 1500 msec
(ungated) or 700 msec (+ a gating delay) to make the actual TR of the gated images approximately equal, a TE of 20 msec in all, and a 256 x 256 image matrix. RESULTS Figure 2 shows a set of coronal images acquired with (a)-(d), and without, (e)-(h), respiratory gating. The liver and kidney can be clearly seen in (c) and (d), along with the stomach and diaphragm. The aorta, vena cava, and vessels branching of these can also be seen in these images. Images (e)-(h) correspond spatially to images (a)-(d), but as they were acquired without gating the anatomical detail of images (a)-(d) is lacking; images (e) and (f) show quite good anatomi-
cal detail as they correspond to slices through positions just above the spine and through the spine, where there is little or no respiratory motion. As the image plane moves up through the body of the rat, respiratory motion increases, exacerbating motion artifacts and degrading image quality, as can be seen in the ungated images. CONCLUSIONS In summary, we present a technique for obtaining good quality MR images of the rat abdomen in vivo. The technique reduces artifacts due to motion sufficiently to allow MRI studies to be performed, thereby providing noninvasive access to animal models of the pathology and physiology of these organs in vivo. Its application to human subjects should be easier than in
Magnetic
900
Resonance
Imaging
Fig.
small animals due to the longer respiratory the larger range of movement.
cycle and
Acknowledgmenfs- It is a pleasure to thank Dr. Herchel Smith for a generous endowment (TAC, LDH) and for a research studentship (NGB), and Clifford Bunch for his technical assistance.
REFERENCES Reeker, F.; Rubben, H.; Bex, A.; Constantinides, C. Morphological changes following ESWL in the rat kidney. Ural. Res. 111229-233; 1989. Brauer, M.; Towner, R.A.; Renaud, 1.; Janzen, E.G.; Foxall, D.L. In vivo proton nuclear magnetic resonance imaging and spectroscopic studies of halocarbon-induced liver damage. Magn. Reson. Med. 9:229-239; 1989. Axel, L.; Summers, R.M.; Kressel, H.Y.; Charles, C. Respiratory effects in two-dimensional Fourier transform MR imaging. Radiology 160:795-801; 1986.
0 Volume
11, Number 6, 1993
2 continued.
4. Thompsen,
C.; Henriksen, 0.; Ring, P. In vivo measurements of relaxation processes in the human liver by MRI. The role of respiratory gating/triggering. Magn. Reson. Imaging 6:431-436; 1988. 5. Aisen, A.M.; Glazer, G.M.; Carson, P.L.; Hearshen, D.O. Motion artifacts in quantitative magnetic resonance imaging. Magn. Reson. Imaging 4:207-213; 1986. 6. Bailes, D.; Gilderdale, D.; Bydder, G.M.; Collins, A.G.; Firmin, D. ROPE a method for reducing respiratory motion artifacts in imaging. J. Comput. Assist. Tomogr. 9: 835-842; 1985. Erlichman, M. Surface/specialty coil devices and gating techniques in magnetic resonance imaging. Health Techno/. Assess. Rep. 3:1-23; 1990. Van Bruggen, N.; Syha, J.; Busza, A.L.; King, M.D.; Stamp, G.W.; Williams, S.R.; Gadian, D.G. Identification of tumor hemorrhage in an animal model using spin echoes and gradient echoes. Magn. Reson. Med. 15:121127; 1990. Ehman, R.L.; McNamara, M.T.; Pallack, M.; Hricak,
Simple device for respiratory gating 0 N.G. BURDETT ET AL. H.; Higgins, C.B. Magnetic resonance imaging with respiratory gating. Techniques and advantages. Am. J. Roentgenol. 143: 1175-l 182; 1984. 10. Cuppen, J.J.M.; Groen, J.; In Den Kleef, J.; Tuithof, H.A. Reduction of artifacts by data post processing. In: Book of abstracts: Fourth Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA: SMRM; 1985:962-963. 11. Buikman, D.; Helzel, T.; Roschcmann, P. The RF coil
901
as a sensitive motion detector. Magn. Reson. Imaging 6:281-289;
1988. 12. Shaff, M.I.; Tarr, R.W.; Partain,
CL.; James, A.E. Computed tomography and magnetic resonance imaging of the acute abdomen. Surg. Clin. North Am. 68: 233-254; 1988. 13. Ballon, D.; Graham, M.C.; Miodownik, S.; Koutcher, J.A. A 64 MHz half-birdcage resonator for clinical imaging. J. Magn. Reson. 90:131-140; 1990.