M7gneric Resonmcelmging. Vol.10.pp.LIB-114, 1992
013%725X02 $5.00 + .oO Copyright 0 1992 Pergamon Press pit
Printed m the USA. All rights reserved.
0 Original Contribution
,AN ESR-CT IMAGING OF THE HEAD OF A LIVING RAT RECEIVING AN ADMINISTRATION OF A NITROXIDE RADICAL SHIN-ICHIISHIDA,* SEIJIMATSUMOTO,* HIDEKATSUYOKOYAMA,* NOFUOMOFU,* HISASHIKUMASHIRO ,* NOBUAKITSUCHIHASHI, -f TATEAKIOGATA, $ MINORU YAMADA,+ MITSUHIROONO, 9 TATSUO KITAJIMA,~ HITOSHIKAMADA,$ AND EKUO YOSHJDA~1 *Department of Neuropsychiatry, Fukushima Medical College, Fukushima 960-12, Japan, tRl Laboratory, Fukushima Medical College, Fukushima 960-12, Japan, $Department of Applied Chemistry, Faculty of Engineering, Yamagata University, Yonezawa 992, Japan, SDepartment of Electronic Engineering, Faculty of Engineering, Yamagata University, Yonezawa 992, Japan, IlDepartment of Information Engineering, Faculty of Engineering, Yamagata University, Yonezawa 992, Japan, and l/Technical and Engineering Division, JEOL Ltd., Akishima 196, Japan. Three-dimensional ESR imaging of a living rat has been performed by an L-band ESR system, which is composed of an L-band ESR spectrometer, a Reid gradient coil, and a data processor. The imaging was carried out by Lauterbur’s method. A nitroxide, 3-carbamoyl-2,2,5,5-tetramethylpyrrolidine-l-oxyl (CarbamoylPROXYL), was used as an imaging agent in saline solution at a concentration of 0.2 M and administered intraperitoneahy to obtain a constant concentration in the head for about an hour. It took about 40 min to obtain one set of ESR-CT images. The cross-sectional images were made, both as coronal and horixontal hnages. In the images of the rat head the nitroxide-rich region was clearly distinguished from the deficient region. The nitroxide-deficient areas corresponded well to the brain of the rat. Keywords: ESR-CT; L-band ESR; Rat head; Nitroxide. INTRODUCTION
Previously we have reported a developments of an L-band ESR system for large biological subjects.6 Using this system we have performed an in vivo measurement of nitroxides in the rat head which received injections of the radicals.’ On the basis of this system we developed an ESR imaging system and obtained two-dimensional images of the distribution of a nitroxide radical injected into mouse legs.8 In the present paper we will describe the three-dimensional ESR imaging of the head of a living rat which received an administration of a stable nitroxide radical.
Free radicalssuch as active oxygens are concerned with
inflammation, ischemic heart disease, carcinogenesis, aging, and pathogenesis and clinical course of a number of neuropsychiatric disorders. Since most of radical species are chemically highly active, they rapidly react with the neighboring molecules and evoke physiological dysfunction in the biological system. Therefore in vivo analysis of the free radicals is important. Moreover it is required to display as images showing the location and the spatial distribution of radicals in the living biological systems. One- or two-dimensional ESR imagings have been made at L-band (usually about 1 GHz), where the subjects were a celery or a mouse tail.1-3 Recently threedimensional imagings of the nitroxide distribution in a sample tube and a living rat tail have been reported.4*5
L-BAND ESR IMAGING SYSTEM
The block diagram of our L-band ESR imaging system is shown in Fig. 1. The system is composed of an
This work is partially supported by Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan. RECEIVED 9/21/90; Accnmnn 6126191.
Address all correspondence to Nobuaki Tsuchihashi, RI Laboratory, Fukushima Medical College, Fukushima 960-12, Japan. 109
110
Magnetic Resonance Imaging 0 Volume 10, Number 1, 1992
phase
shifter
I
AFC in low
divider L-band synthesizer
noise
amp.
RF out
preamp
+
r
iI L-,,,--...-
a 1 I
CPU LSI-11
. . . . . . . ..__
I
spec tro meter
1
~~..~.~..._~~~~~~~_~~~~
Fig. 1. The block diagram of the L-band ESR imaging system equipped with the on-line digital system. This system was operated at 700-800 MHz of the microwave frequency.
L-band ESR spectrometer, a pair of field gradient coils, and a minicomputer. ESR was observed at 700-800 MHz using the L-band ESR spectrometer equipped with a loop-gap resonator having an electric shield in the loop as described previously.’ The loop-gap resonator used at the ESR measurement is of the two-gap type, which is 41 mm in the inner diameter and 10 mm in the axial length. This resonator can afford a measurement of a rat head. A pair of magnetic field gradient coils for X-, Y-, and Z-axes were attached to the surface of the pole pieces of the electromagnet of the L-band ESR spectrometer. The distance between the surfaces of the gradient coils was 79 mm. These coils are consisted of an anti-Hehnholz coil for the Z-axis and the so-called simcoil for the X- and Y-axes. The Z-axis is parallel to the static magnetic field. These coils provide a linear gradient magnetic field of up to 4 mT/cm in the range of 15 mm from the center, In this study, magnetic field gradient of 1 mT/cm was applied. In order to collect the ESR data and to control the system, a minicomputer (DEC:LSI-1 1, USA) was connected to our L-band ESR spectrometer. The magnetic field sweep, the setting up of gradient amplitude and the direction, and the collection of ESR spectral data were automatically performed.
THREE-DIMENSIONAL
IMAGING METHOD
Three-dimensional ESR image is constructed based on the Lauterbur’s method,‘.” known in three-dimensional zeugmatography. The method consists of two steps, and the outline is shown in Fig. 2. The first step is construction of two-dimensional images on several planes from ESR spectral data obtained by changing the direction of the magnetic field gradient. We assume that each plane contains the X-axis and is normal to the YZplane. Let 8i be an angle from the X-axis. If we change 0, at 20” from 0” to 160” on the plane A in Fig. 2, nine ESR spectral data are obtained. Then we can construct a two-dimensional ESR image, GA, on the plane A by the filtered back projection method. Let B be the plane rotated by t3afrom the plane A. By the same procedure made in the plane A, another two-dimensional ESR image, Gu, is obtained. Continuing these procedures by changing 8, at 20” from 0” to 160”, nine two-dimensional images, GA-G,, are obtained. The second step is construction of an image in the slice plane based on these two-dimensional .images. Let the plane M be one slice plane which is parallel to the YZ-plane. The transverse section of one two-dimensional image GA with the
ESR-CT
imaging of living rat head 0 S.
x
&DA
ET AL.
(A)
ta) X t
100mm/200point~
3L
i”
Fig. 3. Example of the treatments of ESR spectral data. (0, = 8, = 80”). (A) The original fit-derivative ESR spectrum (microwave frequency, 710.166 MHz; microwave power, 10 mW; magnetic field modulation width, 0.1 mT; magnetic field gradient, 1 mT/cm; amplitude, 2 * 100; sweep width, 5 mT). (B) The integrated spectrum obtained from the original firstderivative ESR spectral data. (C) The integrated spectrum of the low-field component.
,
0.5mm
lhickness/plans
Fig. 2. Schematic representations of Lauterbur’s method. 8, = The angle of the magnetic field gradient direction from the X-axis. 8, = The angle of the rotation of the projection plane around the X-axis. The values of 8, and f12were changed stepwise by each 20” in this study.
plane M, gM, is also regarded as one-dimensional profile of ESR spectra. So, based on these nine data,
gW g,, we can construct a two-dimensional ESRimage on the plane M also by the filtered back projection. The two-dimensional image constructed in the above-mentioned way is termed here an ESR-CT image at the plane M. As ESR spectral data recorded in the first step are first-derivative spectra, they are integrated first. As is generally known, the random noises were eliminated by this treatment. Though ESR spectra of nitroxide radicals such as Carbamoyl-PROXYL (see next section) have three hyperfine structure lines due to N-14 nucleus (Z = I), only one of these lines may be sufficient for image construction.” So, two line in the higher field were eliminated here. These processes are shown in Fig. 3. The data based on the low-field component
112
Magnetic Resonance Imaging 0 Volume 10, Number 1, 1992
MODULATION
COIL
Fig. 4. Photograph of the positioning of the rat head in the shield case (upper) and its schematic representation (lower).
from each field gradient direction were subjected to the image reconstruction, without a deconvolution of the line shape in this study. In the following example, it took about 40 min to measure 81 ESR spectral data required for construction of an ESR-CT image. Data processing was carried out by a superminicomputer (MVlOOOOSX, Nippon Data General, Japan), and all images were displayed on a personal computer (PC9801, NEC, Japan). Under 1 mT/cm gradient, data on lOO-mm (= 10 mT) line for X, Y, and Z directions were assigned to 200 data points,
namely one point represents 0.5 mm. The data from 50 mm * 50 mm (100 points * 100 points) in the central region were chosen and subjected to the image reconstruction. The region is enough to include the resonator (41 mm I+) with the rat head. When we display an ESR-CT image at the plane A4 practicahy, it will be better to average it by two ESR-CT images at adjacent planes to M, because of measurement errors of ESR spectra and round off errors in data processing. In this paper, an ESR-CT image at the plane M was averaged by images at two adjacent planes, one
ESR-CTimaging of living rat head 0 S.
113
ISHIDA ET AL.
RESONATOR
klomrnA Cerebrum urn
-5
-b
(mm)
Incisor
tooth
t
t
+10 D
t
t
0 E
F
t -10
G
(mm)
H
Fig. 5. ESR-CT of horizontal (A, B, C) and coronal (E, F, G and D, H) sections and the schema of the sagittal section of a rat head. The rat received the intraperitoneal administration of 5 ml of 0.2 M Carbamoyl-PROXYL under pentobarbital anesthesia, The images (D and H, both 5 mm from the edge of the resonator) due to the ESR signals detected by the external microwave magnetic field. Thickness of slice plane of the ESR-CT was 1.5 mm. For measurement conditions, see Fig. 3.
point away from the plane M. This means that an ESR-CT image in the following was considered as an image on the slice plane with 1.5 mm thick. Data of one point were visualized by four matrices in the color display. Image signals were scaled by eight levels, each level being 12.5% of the maximum signal intensity. The signals lower than 12.5% of the maximum signal level were regarded as noise. THREE-DIMENSIONAL IMAGING OF THE RAT HEAD
Male Wistar rats weighing about 120 g were used. Nitroxide of 3-carbamoyl-2,2,5,5_tetramethyl-
pyrrolidine- 1-oxyl (Carbamoyl-PROXYL) was used as an imaging agent in saline solution at a concentration of 0.2 M. The rat head was placed in the loop-gap resonator under pentobarbital anesthesia. The rat received intraperitoneal administration of 5 ml of the Carbamoyl-PROXYL solution to obtain a constant concentration in the head for about an hour. After these treatment the rat was set in the imaging system and ESR measurement was started under the condition of the resonant frequency 710 MHz, the microwave power 10 mW, and the modulation width 0.1 mT. Figure 4 shows the positioning of the rat head in the shield case and in the Cartesian coordinates. Figure 5 shows a set of the ESR-CT images of the rat which re-
114
Magnetic Resonance Imaging 0 Volume 10, Number 1, 1992
ceived the administration of Carbamoyl-PROXYL. The ESR-CT of the coronal (XZ-plane) and the horizontal (YZ-plane) sections was performed and is displayed in the figure, together with the schema of the sagittal sections of a rat head. Thickness of the slice plane was 1.5 mm. Comparing the images with the schema, one can find that the nitroxide-rich region corresponds to the position where many blood vessels and/or many blood pools locate and was clearly distinguished from the deficient region. The nitroxide-deficient areas corresponded well to the brain of the rat. Since the brain holds the blood brain barrier, Carbamoyl-PROXYL may be retained here, and its concentration in the brain is quite low, which results in weak ESR signals, while its high concentration in the blood results in strong ESR signals. The nitroxide radicals are also present in the rat head outside the resonator, because the radicals are circulating in the rat body after the intraperitoneal injection. The horizontal images outside the resonator correspond to the signals which originated from the nitroxide radicals detected by the external microwave magnetic field outside the resonator. The imaging outside the resonator, which are indicated by the arrows (D and H; both 5 mm from the edge of the resonator), shows that the external microwave magnetic field presents at the regions. Recently three-dimensional imaging of the nitroxide distribution in a sample tube (9 mm in diameter) and in a living rat tail have been reported.4’5 A large subject such as a rat head induces a decrease of the loaded Q-factor. The fluctuations of the resonant frequency by breathing, the heartbeat, and the blood flow induce noise signals. Thus we have some difficulties to overcome in performing in vivo measurements and subsequent ESR imaging. The best matching was obtained by moving the coupling coil and adjusting the three-stub tuner.’ Figure 1 indicates this status of the resonant conditions. Thus we obtained an ESR spectrum with a good signal-to-noise ratio eliminating the influence from the breathing of a living rat in the course of the ESR observation, though in the imaging study the signal-to-noise ratio of the ESR spectrum was reduced by magnetic field gradient which induces broadening of the ESR spectrum. The ESR-CT is a straightforward way to obtain informations on the distribution of free radicals in large subjects, and to our knowledge this is the first report of
ESR-CT images for a living rat head. Another imaging technique for free radicals based on the Overhauser effect has been reported,‘*-I4 and it shows promise as well as the ESR-CT. CONCLUSION We developed an L-band ESR imaging system. Using this system we have successfully made ESR-CT images of the head of the rat which received the intraperitoneal administration of Carbamoyl-PROXYL solution. The images obtained here gave good information on the distribution of exogenous nitroxide radicals in the rat head. We believe our system has been well developed as an L-band ESR imaging system and it would be available to analyze the spatial distribution of the nitroxide radicals in the biological objects. REFERENCES 1. Berliner, L.J.; Fujii, H. Science 227517-519; 1985. 2. Fujii, H.; Berliner, L.J. Magn. Resort. Med. 2~275-282; 1985.
3. Berliner, L.J.; Fujii, H.; Wan, X.; Lukiewicz, S.J. Magn. Reson. Med. 4:380-384; 1987.
4. Colacicchi, S.; Indovina, P.L.; Mono, F.; Sotgiu, A. J. Phys. E. Sci. Instrum. 21:910-913; 1988.
5. Alecci, M.; Colacicchi, S.; Indovina, P.L.; Mono, F.; 6. 7. 8.
9. 10. 11. 12. 13.
Pavone, P.; Sotgiu, A. Magn. Reson. Imag. 8:59-63; 1990. Ono, M.; Ogata, T.; Hsieh, KC.; Suzuki, M.; Yoshida, E.; Kamada, H. Chem. Leti. 491-494; 1986. Ishida, S.; Kumashiro, H.; Tsuchihashi, N.; Ogata, T.; Ono, M.; Kamada, H.; Yoshida, E. Phys. Med. Biol. 34: 1317-1323; 1989. Ishida, S.; Matsumoto, S.; Yokoyama, H.; Mori, N.; Kumashiro, H.; Tsuchihashi, N.; Ogata, T.; Kitajima, T.; Ono, M.; Kamada, H.; Yoshida, E. Jpn. J. Magn. Reson. Med. 10:21-27; 1990. Lauterbur, P.C. Nature 242:190-191; 1973. Lauterbur, P.C.; Lai, C.M. IEEE Trans. Nucl. Sci. NS27:1227-1231; 1980. Ohno, K. Appl. Spectrosc. Rev. 22:1-56; 1986. Grucker, D. Magn. Reson. Med. 14:14O-147; 1990. Lurie, D.J.; Bussell, D.M.; Bell, L.H.; Mallard, J.R. J. Magn. Reson. 76:36&370;
1988.
14. Lurie, D.J.; Hutchison, J.M.S.; Bell, L.H.; Nicholson, I.; Bussell, D.M.; Mallard, J.R. J. Magn. Reson. 84:43 l437; 1989.
013%725X02 $5.00 + .oO Copyright 0 1992 Pergamon Press pit
Printed m the USA. All rights reserved.
0 Original Contribution
,AN ESR-CT IMAGING OF THE HEAD OF A LIVING RAT RECEIVING AN ADMINISTRATION OF A NITROXIDE RADICAL SHIN-ICHIISHIDA,* SEIJIMATSUMOTO,* HIDEKATSUYOKOYAMA,* NOFUOMOFU,* HISASHIKUMASHIRO ,* NOBUAKITSUCHIHASHI, -f TATEAKIOGATA, $ MINORU YAMADA,+ MITSUHIROONO, 9 TATSUO KITAJIMA,~ HITOSHIKAMADA,$ AND EKUO YOSHJDA~1 *Department of Neuropsychiatry, Fukushima Medical College, Fukushima 960-12, Japan, tRl Laboratory, Fukushima Medical College, Fukushima 960-12, Japan, $Department of Applied Chemistry, Faculty of Engineering, Yamagata University, Yonezawa 992, Japan, SDepartment of Electronic Engineering, Faculty of Engineering, Yamagata University, Yonezawa 992, Japan, IlDepartment of Information Engineering, Faculty of Engineering, Yamagata University, Yonezawa 992, Japan, and l/Technical and Engineering Division, JEOL Ltd., Akishima 196, Japan. Three-dimensional ESR imaging of a living rat has been performed by an L-band ESR system, which is composed of an L-band ESR spectrometer, a Reid gradient coil, and a data processor. The imaging was carried out by Lauterbur’s method. A nitroxide, 3-carbamoyl-2,2,5,5-tetramethylpyrrolidine-l-oxyl (CarbamoylPROXYL), was used as an imaging agent in saline solution at a concentration of 0.2 M and administered intraperitoneahy to obtain a constant concentration in the head for about an hour. It took about 40 min to obtain one set of ESR-CT images. The cross-sectional images were made, both as coronal and horixontal hnages. In the images of the rat head the nitroxide-rich region was clearly distinguished from the deficient region. The nitroxide-deficient areas corresponded well to the brain of the rat. Keywords: ESR-CT; L-band ESR; Rat head; Nitroxide. INTRODUCTION
Previously we have reported a developments of an L-band ESR system for large biological subjects.6 Using this system we have performed an in vivo measurement of nitroxides in the rat head which received injections of the radicals.’ On the basis of this system we developed an ESR imaging system and obtained two-dimensional images of the distribution of a nitroxide radical injected into mouse legs.8 In the present paper we will describe the three-dimensional ESR imaging of the head of a living rat which received an administration of a stable nitroxide radical.
Free radicalssuch as active oxygens are concerned with
inflammation, ischemic heart disease, carcinogenesis, aging, and pathogenesis and clinical course of a number of neuropsychiatric disorders. Since most of radical species are chemically highly active, they rapidly react with the neighboring molecules and evoke physiological dysfunction in the biological system. Therefore in vivo analysis of the free radicals is important. Moreover it is required to display as images showing the location and the spatial distribution of radicals in the living biological systems. One- or two-dimensional ESR imagings have been made at L-band (usually about 1 GHz), where the subjects were a celery or a mouse tail.1-3 Recently threedimensional imagings of the nitroxide distribution in a sample tube and a living rat tail have been reported.4*5
L-BAND ESR IMAGING SYSTEM
The block diagram of our L-band ESR imaging system is shown in Fig. 1. The system is composed of an
This work is partially supported by Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan. RECEIVED 9/21/90; Accnmnn 6126191.
Address all correspondence to Nobuaki Tsuchihashi, RI Laboratory, Fukushima Medical College, Fukushima 960-12, Japan. 109
110
Magnetic Resonance Imaging 0 Volume 10, Number 1, 1992
phase
shifter
I
AFC in low
divider L-band synthesizer
noise
amp.
RF out
preamp
+
r
iI L-,,,--...-
a 1 I
CPU LSI-11
. . . . . . . ..__
I
spec tro meter
1
~~..~.~..._~~~~~~~_~~~~
Fig. 1. The block diagram of the L-band ESR imaging system equipped with the on-line digital system. This system was operated at 700-800 MHz of the microwave frequency.
L-band ESR spectrometer, a pair of field gradient coils, and a minicomputer. ESR was observed at 700-800 MHz using the L-band ESR spectrometer equipped with a loop-gap resonator having an electric shield in the loop as described previously.’ The loop-gap resonator used at the ESR measurement is of the two-gap type, which is 41 mm in the inner diameter and 10 mm in the axial length. This resonator can afford a measurement of a rat head. A pair of magnetic field gradient coils for X-, Y-, and Z-axes were attached to the surface of the pole pieces of the electromagnet of the L-band ESR spectrometer. The distance between the surfaces of the gradient coils was 79 mm. These coils are consisted of an anti-Hehnholz coil for the Z-axis and the so-called simcoil for the X- and Y-axes. The Z-axis is parallel to the static magnetic field. These coils provide a linear gradient magnetic field of up to 4 mT/cm in the range of 15 mm from the center, In this study, magnetic field gradient of 1 mT/cm was applied. In order to collect the ESR data and to control the system, a minicomputer (DEC:LSI-1 1, USA) was connected to our L-band ESR spectrometer. The magnetic field sweep, the setting up of gradient amplitude and the direction, and the collection of ESR spectral data were automatically performed.
THREE-DIMENSIONAL
IMAGING METHOD
Three-dimensional ESR image is constructed based on the Lauterbur’s method,‘.” known in three-dimensional zeugmatography. The method consists of two steps, and the outline is shown in Fig. 2. The first step is construction of two-dimensional images on several planes from ESR spectral data obtained by changing the direction of the magnetic field gradient. We assume that each plane contains the X-axis and is normal to the YZplane. Let 8i be an angle from the X-axis. If we change 0, at 20” from 0” to 160” on the plane A in Fig. 2, nine ESR spectral data are obtained. Then we can construct a two-dimensional ESR image, GA, on the plane A by the filtered back projection method. Let B be the plane rotated by t3afrom the plane A. By the same procedure made in the plane A, another two-dimensional ESR image, Gu, is obtained. Continuing these procedures by changing 8, at 20” from 0” to 160”, nine two-dimensional images, GA-G,, are obtained. The second step is construction of an image in the slice plane based on these two-dimensional .images. Let the plane M be one slice plane which is parallel to the YZ-plane. The transverse section of one two-dimensional image GA with the
ESR-CT
imaging of living rat head 0 S.
x
&DA
ET AL.
(A)
ta) X t
100mm/200point~
3L
i”
Fig. 3. Example of the treatments of ESR spectral data. (0, = 8, = 80”). (A) The original fit-derivative ESR spectrum (microwave frequency, 710.166 MHz; microwave power, 10 mW; magnetic field modulation width, 0.1 mT; magnetic field gradient, 1 mT/cm; amplitude, 2 * 100; sweep width, 5 mT). (B) The integrated spectrum obtained from the original firstderivative ESR spectral data. (C) The integrated spectrum of the low-field component.
,
0.5mm
lhickness/plans
Fig. 2. Schematic representations of Lauterbur’s method. 8, = The angle of the magnetic field gradient direction from the X-axis. 8, = The angle of the rotation of the projection plane around the X-axis. The values of 8, and f12were changed stepwise by each 20” in this study.
plane M, gM, is also regarded as one-dimensional profile of ESR spectra. So, based on these nine data,
gW g,, we can construct a two-dimensional ESRimage on the plane M also by the filtered back projection. The two-dimensional image constructed in the above-mentioned way is termed here an ESR-CT image at the plane M. As ESR spectral data recorded in the first step are first-derivative spectra, they are integrated first. As is generally known, the random noises were eliminated by this treatment. Though ESR spectra of nitroxide radicals such as Carbamoyl-PROXYL (see next section) have three hyperfine structure lines due to N-14 nucleus (Z = I), only one of these lines may be sufficient for image construction.” So, two line in the higher field were eliminated here. These processes are shown in Fig. 3. The data based on the low-field component
112
Magnetic Resonance Imaging 0 Volume 10, Number 1, 1992
MODULATION
COIL
Fig. 4. Photograph of the positioning of the rat head in the shield case (upper) and its schematic representation (lower).
from each field gradient direction were subjected to the image reconstruction, without a deconvolution of the line shape in this study. In the following example, it took about 40 min to measure 81 ESR spectral data required for construction of an ESR-CT image. Data processing was carried out by a superminicomputer (MVlOOOOSX, Nippon Data General, Japan), and all images were displayed on a personal computer (PC9801, NEC, Japan). Under 1 mT/cm gradient, data on lOO-mm (= 10 mT) line for X, Y, and Z directions were assigned to 200 data points,
namely one point represents 0.5 mm. The data from 50 mm * 50 mm (100 points * 100 points) in the central region were chosen and subjected to the image reconstruction. The region is enough to include the resonator (41 mm I+) with the rat head. When we display an ESR-CT image at the plane A4 practicahy, it will be better to average it by two ESR-CT images at adjacent planes to M, because of measurement errors of ESR spectra and round off errors in data processing. In this paper, an ESR-CT image at the plane M was averaged by images at two adjacent planes, one
ESR-CTimaging of living rat head 0 S.
113
ISHIDA ET AL.
RESONATOR
klomrnA Cerebrum urn
-5
-b
(mm)
Incisor
tooth
t
t
+10 D
t
t
0 E
F
t -10
G
(mm)
H
Fig. 5. ESR-CT of horizontal (A, B, C) and coronal (E, F, G and D, H) sections and the schema of the sagittal section of a rat head. The rat received the intraperitoneal administration of 5 ml of 0.2 M Carbamoyl-PROXYL under pentobarbital anesthesia, The images (D and H, both 5 mm from the edge of the resonator) due to the ESR signals detected by the external microwave magnetic field. Thickness of slice plane of the ESR-CT was 1.5 mm. For measurement conditions, see Fig. 3.
point away from the plane M. This means that an ESR-CT image in the following was considered as an image on the slice plane with 1.5 mm thick. Data of one point were visualized by four matrices in the color display. Image signals were scaled by eight levels, each level being 12.5% of the maximum signal intensity. The signals lower than 12.5% of the maximum signal level were regarded as noise. THREE-DIMENSIONAL IMAGING OF THE RAT HEAD
Male Wistar rats weighing about 120 g were used. Nitroxide of 3-carbamoyl-2,2,5,5_tetramethyl-
pyrrolidine- 1-oxyl (Carbamoyl-PROXYL) was used as an imaging agent in saline solution at a concentration of 0.2 M. The rat head was placed in the loop-gap resonator under pentobarbital anesthesia. The rat received intraperitoneal administration of 5 ml of the Carbamoyl-PROXYL solution to obtain a constant concentration in the head for about an hour. After these treatment the rat was set in the imaging system and ESR measurement was started under the condition of the resonant frequency 710 MHz, the microwave power 10 mW, and the modulation width 0.1 mT. Figure 4 shows the positioning of the rat head in the shield case and in the Cartesian coordinates. Figure 5 shows a set of the ESR-CT images of the rat which re-
114
Magnetic Resonance Imaging 0 Volume 10, Number 1, 1992
ceived the administration of Carbamoyl-PROXYL. The ESR-CT of the coronal (XZ-plane) and the horizontal (YZ-plane) sections was performed and is displayed in the figure, together with the schema of the sagittal sections of a rat head. Thickness of the slice plane was 1.5 mm. Comparing the images with the schema, one can find that the nitroxide-rich region corresponds to the position where many blood vessels and/or many blood pools locate and was clearly distinguished from the deficient region. The nitroxide-deficient areas corresponded well to the brain of the rat. Since the brain holds the blood brain barrier, Carbamoyl-PROXYL may be retained here, and its concentration in the brain is quite low, which results in weak ESR signals, while its high concentration in the blood results in strong ESR signals. The nitroxide radicals are also present in the rat head outside the resonator, because the radicals are circulating in the rat body after the intraperitoneal injection. The horizontal images outside the resonator correspond to the signals which originated from the nitroxide radicals detected by the external microwave magnetic field outside the resonator. The imaging outside the resonator, which are indicated by the arrows (D and H; both 5 mm from the edge of the resonator), shows that the external microwave magnetic field presents at the regions. Recently three-dimensional imaging of the nitroxide distribution in a sample tube (9 mm in diameter) and in a living rat tail have been reported.4’5 A large subject such as a rat head induces a decrease of the loaded Q-factor. The fluctuations of the resonant frequency by breathing, the heartbeat, and the blood flow induce noise signals. Thus we have some difficulties to overcome in performing in vivo measurements and subsequent ESR imaging. The best matching was obtained by moving the coupling coil and adjusting the three-stub tuner.’ Figure 1 indicates this status of the resonant conditions. Thus we obtained an ESR spectrum with a good signal-to-noise ratio eliminating the influence from the breathing of a living rat in the course of the ESR observation, though in the imaging study the signal-to-noise ratio of the ESR spectrum was reduced by magnetic field gradient which induces broadening of the ESR spectrum. The ESR-CT is a straightforward way to obtain informations on the distribution of free radicals in large subjects, and to our knowledge this is the first report of
ESR-CT images for a living rat head. Another imaging technique for free radicals based on the Overhauser effect has been reported,‘*-I4 and it shows promise as well as the ESR-CT. CONCLUSION We developed an L-band ESR imaging system. Using this system we have successfully made ESR-CT images of the head of the rat which received the intraperitoneal administration of Carbamoyl-PROXYL solution. The images obtained here gave good information on the distribution of exogenous nitroxide radicals in the rat head. We believe our system has been well developed as an L-band ESR imaging system and it would be available to analyze the spatial distribution of the nitroxide radicals in the biological objects. REFERENCES 1. Berliner, L.J.; Fujii, H. Science 227517-519; 1985. 2. Fujii, H.; Berliner, L.J. Magn. Resort. Med. 2~275-282; 1985.
3. Berliner, L.J.; Fujii, H.; Wan, X.; Lukiewicz, S.J. Magn. Reson. Med. 4:380-384; 1987.
4. Colacicchi, S.; Indovina, P.L.; Mono, F.; Sotgiu, A. J. Phys. E. Sci. Instrum. 21:910-913; 1988.
5. Alecci, M.; Colacicchi, S.; Indovina, P.L.; Mono, F.; 6. 7. 8.
9. 10. 11. 12. 13.
Pavone, P.; Sotgiu, A. Magn. Reson. Imag. 8:59-63; 1990. Ono, M.; Ogata, T.; Hsieh, KC.; Suzuki, M.; Yoshida, E.; Kamada, H. Chem. Leti. 491-494; 1986. Ishida, S.; Kumashiro, H.; Tsuchihashi, N.; Ogata, T.; Ono, M.; Kamada, H.; Yoshida, E. Phys. Med. Biol. 34: 1317-1323; 1989. Ishida, S.; Matsumoto, S.; Yokoyama, H.; Mori, N.; Kumashiro, H.; Tsuchihashi, N.; Ogata, T.; Kitajima, T.; Ono, M.; Kamada, H.; Yoshida, E. Jpn. J. Magn. Reson. Med. 10:21-27; 1990. Lauterbur, P.C. Nature 242:190-191; 1973. Lauterbur, P.C.; Lai, C.M. IEEE Trans. Nucl. Sci. NS27:1227-1231; 1980. Ohno, K. Appl. Spectrosc. Rev. 22:1-56; 1986. Grucker, D. Magn. Reson. Med. 14:14O-147; 1990. Lurie, D.J.; Bussell, D.M.; Bell, L.H.; Mallard, J.R. J. Magn. Reson. 76:36&370;
1988.
14. Lurie, D.J.; Hutchison, J.M.S.; Bell, L.H.; Nicholson, I.; Bussell, D.M.; Mallard, J.R. J. Magn. Reson. 84:43 l437; 1989.
