224
Brain Research, 574 (1992) 224-228 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$05.1X)
BRES 17509
Quantitation of ventricular size in normal and spontaneously hypertensive rats by magnetic resonance imaging Peter Bendel I and Raya Eilam 2 Departments of 1Chemical Physics and 2Neurobiology, The Weizmann Institute of Science, Rehovot (Israel) (Accepted 22 October 1991) Key words: Magnetic resonance imaging; Brain ventricular dilation; Spontaneously hypertensive rat; Wystar-Kyoto rat
Magnetic resonance imaging (MRI) performed at high field (4.7 Tesla), and high spatial resolution (0.6 mm slice thickness, 0.18 mm inplane) enabled noninvasive quantitative measurement of the ventricular vol. in live rats. Comparing the results for 15 male Wystar-Kyoto (WKY) rats, aged 2.5-10 months, with those from 17 spontaneously hypertensive rats (SHR), clearly confirmed the previously reported elevated ventricular vols. in the SHR strain. A significant difference in ventrieular vol. between the two strains was detected above the age of 3 months. For mature animals above the age of 6 months the mean vol. in the SHR strain was elevated by about a factor of two compared to the WKY control animals. INTRODUCTION The spontaneously hypertensive rat (SHR) strain has been used extensively as an animal model for essential hypertension in the human. Since the S H R was derived by selective inbreeding from the W K Y strain 18, W K Y are usually considered to be the appropriate normotensive controls for the SHR. These hypertensive rats are characterized by a broad spectrum of anatomical, histological, as well as neuroendocrine deviations 7'1°'14. It was recently reported that brain ventricular size of the S H R is significantly elevated in comparison to ageand sex-matched W K Y rats. The ventricular dilation is associated with a loss of brain tissue 2° and reduced brain weight and vol. 19, which may underlie some of the behavioral abnormalities of SHR. It was found that ventricular enlargement develops abruptly between 4 and 8 weeks of age, thereafter progressing at a slower rate. This pathology was found to occur in the absence of visible obstruction within the ventricular space, and does not appear to result from permeability changes of the blood-brain barrier. The conventional methodology for assessing ventricular size, which led to the findings mentioned above 19'2° involved sacrificing the animal, followed by perfusion and cutting of brain sections. Magnetic resonance imaging (MRI) has the potential for completely noninvasive examination of the brain anatomy. It therefore enables us to follow the development of anatomical pathology in the same animal, which could be correlated with con-
comittant physiological and behavioral alterations. The aim of our study was to develop M R I methods which would permit quantitative noninvasive measurement of brain ventricular vols. and validate those methods. M R I is a powerful radiological diagnostic method for humans, particularly in the central nervous system 6. The spatial resolution typically achieved in medical imaging (slice thickness of the order of 5 ram, in-plane resolution of the order of 1 mm) is not suitable for detecting the fine anatomical details, in particular the ventricular structure, within the rat brain. U p to now, M R images of rat brain 16, or attempts to quantitate enlarged rat brain ventricles by MR122 did not achieve the resolution necessary for measuring ventriclular size in animals with normal or moderately enlarged ventricles. We succeeded in attaining improved resolution, enabling clear demonstration of the ventricles. The key factors which permitted this were increased signal-to-noise ratio (S/N) achieved due to the high magnetic field, radio frequency (rf) and gradient hardware which is matched to the size of small samples, and by applying imaging protocols which maximize the contrast between cerebro-spinal fluid (CSF) and surrounding brain tissue.
MATERIALS AND METHODS The MR images presented in this report were obtained on a Bruker Biospec 47/30 NMR system, operating at 4.7 Tesla field strength (resonance frequency for protons: 200 MHz) with a 30 cm diameter horizontal access bore. The S/N was maximized by using a home-built rf coil of 5 cm diameter which closely fits over the
Correspondence: P. Bendel, Department of Chemical Physics, The Weizmann Institute of Science, Rehovot 76100, Israel.
225 Fig. 1. Example of a sagittal localization image, obtained in 1.3 minutes with repetition time (TR) of 600 ms and echo time (TE) of 17 ms. The slice thickness is 2 mm. The coronal slices for the study which will be used for the ventricle size measurements are selected on this image by interactive software.
head of the animal. After initial evaluation with different imaging protocols, using varying parameters, we settled on the following routine procedure for obtaining the MR images: The rats were anesthetized with pentobarbital (about 40 mg/kg i.p.). A rapid orientation image was obtained (in about 2 min of scan time) in a sagittal plane, and the coronal slices for the subsequent study were selected on this orientation image, perpendicular to the anteriorposterior axis of the brain (Fig. 1). Twelve to 18 slices were selected to cover the region between the septal caudate-putamen nuclei and the substantia nigra nucleus. Slice thickness was 0.6 mm in most studies, with an interslice gap of 0.09 mm. (Some of the studies were conducted with slice thickness 0.87 mm and gaps of 0.13
~i
•
i ~¸
~
Fig. 2. Consecutive representative sfices from T2-weighted magnetic resonance imaging (MRI) scans (TR = 3.3 s, TE = 68 ms) for a spontaneously hypertensive rat (SHR) (right) and WKY (left) rat. The order of advance from posterior to anterior is left to right and top to bottom. The small square for each slice represents an area of 2.25 × 2.25 cm 2. Slice thickness is 0.6 ram, and the center-to-center distance between slices is 0.69 ram. The WKY rat is listed #10, and the SHR #13 in Table I.
226 80
v
60
~,
4o
~
20
IT/
~t
> 0
2.5
3-5 Age (months)
lated by the computer and displayed. Total ventricular vols. were determined by adding the areas of all the ventricular structures on the monitored slices, multiplied by the distance between adjacent slice centers. This post-processing procedure took an additional half hour per study. RESULTS T h e results for c o m p u t e d v e n t r i c u l a r vols. f r o m a total of 27 a n i m a l studies are t a b u l a t e d in Table I and graphically displayed in Fig. 3. T h e values w e r e c o m p u t e d as i n d i c a t e d in the p r e c e d i n g M a t e r i a l s and M e t h -
6-8
ods section. In 10 of t h e studies, the v e n t r i c u l a r vol. was
Fig. 3. Mean ventricular vol. vs age, summarizing the results in Table I, for WKY (empty boxes) and SHR (shaded boxes) rats. Significant differences between groups are shown by asterixes.** P < 0.05.
c o m p u t e d twice by the s a m e o b s e r v e r (P.B.) to assess i n t r a r a t e r reliability (r = 0.989). T h e s a m e 10 studies w e r e also a n a l y z e d by a s e c o n d o b s e r v e r ( R . E . ) , yielding r = 0.979 for the c o r r e l a t i o n b e t w e e n the values of
mm.) The field-of-view (FOV) was 4.5 cm on a 256 x 256 matrix. Images were acquired with a standard 2DFT spin-echo protocol, using a repetition time (TR) of 3.3 s, a single echo with echo time (TE) of 68 ms, and 2 averages per cycle. The total imaging time for the coronal images was therefore 29 rain and the total experimental study time for each animal (including positioning, calibrations and orientation image) was about 45 rain. The images acquired under the above conditions are T2-weighted, with the CSF, which has a relatively long T2 relaxation time, appearing at much higher intensity than the surrounding brain tissue, thus clearly delineating the outline of the ventricular structure. Fig. 2 shows the results of such T2-weighted coronal slices, displaying the images from representative studies for an SHR and a WKY rat. The standard image-display software on the instrument allows the interactive delineation of arbitrary regions-of-interest (ROI) on the images (by traekbaU). The areas of these ROl's were then calcu-
o b s e r v e r II and the m e a n values of o b s e r v e r I. Considering the scatter in all t h e s e m e a s u r e m e n t s (and including several i m a g i n g e x p e r i m e n t s r e p e a t e d on the s a m e a n i m a l on the s a m e day), a r o o t - m e a n - s q u a r e e r r o r of _+ 5 m m 3 for the r e p o r t e d v e n t r i c u l a r vols. was calculated. T h e m o s t significant p o t e n t i a l e r r o r source for the absolute accuracy of t h e results is the finite c e n t e r - t o - c e n ter distance (of 0.69 m m ) b e t w e e n a d j a c e n t slices. If the v e n t r i c u l a r structure c h a n g e s significantly o v e r this scale, only an a v e r a g e structure will p r o j e c t o n t o the two-dim e n s i o n a l i m a g e in e a c h slice (the so-called 'partial volu m e ' effect), which m a y cause errors in the c o m p u t e d
TABLE I Experimental results for ventricular volumes WKY #
Age (months)
Volume a (mm 3)
Blood pressure
SHR #
Age (months)
Volume ~ (ram 3)
Blood pressure
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
2.5 2.5 2.5 3 3 3 4 4 5 5 6 7 8 8 10
19 13 14 26 29 25 36 26 24 31 29 35 36 36 30
110 115 115 110 130 130
1 2 3 4 5 6 7 8 9 9a 10 11 12 13 14 14a 15 16 17 18 19
2 2 2.5 2.5 2.5 3 3 5 5 6 6 6 6 6 6 7 7.5 7.5 7.5 7.5 13
8 8 16 16 14 38 39 37 36 52 68 68 58 85 54 61 60 51 60 26 56
120 115 130 155 155 185 165 200 200 200 200 205
120 120 135
a Experimental uncertainty: + 5 mm 3. SHR # 9a and 14a are follow-up studies of the same animals as in 9 and 14.
160 220 240 240 235 185
227
Fig. 6. Comparison between a spin density weighted (left) and T2weighted image (right). The images are from a double-echo scan, with TR = 3.3 s, and echo times as indicated.
Fig. 4. Comparison between 0.6 mm thick MRI slices from the live animal (left) and brain sections (thickness: 0.1 mm) obtained sacrificing the same animal (SHR #13), followed by perfusion. Bar = 0.5 cm.
vol. It is difficult to quantitate this possible error in accuracy, but clearly it should be become worse with increasing slice thickness and interslice gap. We therefore conducted some experiments at an additional different setting of 0.87 m m slice thickness and 0.13 m m gap, but found no significant change in the measured ventricular vols., indicating that if a systematic error exists, it should be within the limits of the exPerimental uncertainty. In mature W K Y rats there was no significant variation of ventricular vol. with age. We found the mean vols. to be 28.1 + 1.5 m m 3 (n = 7) for animals aged 3-5 months, and 33.2 + 3.8 mm 3 for 6 - 8 months old rats. For the monitored S H R rats an increase in ventricular vol. was clearly discernible, up to the age of about 7 months. The mean values were 37.5 + 6.6 m m 3 for rats of age 3-5
Fig. 5. Comparison between a /'1- (left) and T2-weighted scan (right). Slice thickness is 0.6 mm, and TRZI'E as indicated. The square for each image is 2.25 x 2.25 cm2. Notice the faint dark appearance of the ventricles on the left image, vs the strong positive contrast between the ventricles and the brain on the right. The bright appearance of the ventricles is due to the long T2 values of the CSE
months (n = 4), and 58.4 + 4.3 m m 3 for ages 6 - 7 months (n = 11). This trend was also reflected in the resuits obtained from two animals in which follow-up studies were conducted a month after the original study. However, the results for two individual studies (26 m m 3 and 85 m m 3) diversed widely from the rest in their age range, indicating a possible strong individual variation of ventricular size for S H R rats in the same age group. A significant difference between the hypertensive and normotensive groups (see Fig. 3) was found in the adult groups (above 3 months), but not between the young animals. Perfused brain sections of W K Y rats and SHR's were compared with the corresponding M R I slices to qualitatively demonstrate the agreement between the M R images and the anatomy as reflected on the perfused sections (Fig. 4). DISCUSSION The crucial factors which enabled the use of M R to image the ventricles in rat brains were maximal S/N, as well as imaging protocol parameters that have fine enough resolution (in terms of slice thickness and FOV) to visualize the details of the ventricular structure and create adequate contrast between CSF and surrounding brain tissue. The S/N necessary to produce 0.6 m m thick slices and 0.18 m m 2 in-plane pixel size could only be achieved at field strengths which are considered high for medical imaging (4.7 Tesla in our case, whereas wholebody medical imagers reach 2 Tesla at most). Close-fitting coils, which maximize the filling factor for the brain, are also important. Finally, the optimal parameters (TR, TE) for the imaging protocol should be chosen with care. A t lower magnetic fields, such as those used for medical imaging, CSF is usually visualized easily by its low signal intensity on Tl-weighted scans. However, at higher fields the 7"1 values for brain gray and white mat-
228 ter increase considerably 5, much m o r e than for C S E A s a result, the contrast on Tl-weighted brain images at high fields is very p o o r , as d e m o n s t r a t e d in Fig. 5, which compares T 1- and T2- weighted scans of the same slice. On spin-density weighted scans, the C S F to brain contrast is also not as g o o d as in T2-weighted images. This is shown in Fig. 6, which displays the first and second echo from a double echo sequence. One of the well known p h e n o m e n a that characterizes S H R ' s is a gradual d e v e l o p m e n t of hypertension as a function of age 17. During the m a t u r a t i o n process, significant loss of brain weight and vol., as well as reduction in cell cross-section areas, c o m p a r e d with normotensive ( W K Y ) rats has also been r e p o r t e d 7'15. Interestingly, S H R were found to display brain ventricular enlargement during the same p e r i o d 2°. This pathology could not have been the direct result of peripheral elevation of blood pressure, since chronic reduction of blood pressure in S H R by capnopnil t r e a t m e n t failed to attenuate the dilation. By the same token, experimentally produced hypertension did not lead to increase in brain ventricular size in W K Y rats 19. In the present study we also found a d e v e l o p m e n t of the cerebral ventricular enlargement in S H R rats aged
REFERENCES 1 Albert, M., Naeser, M.A., Levine, H.L. and Garvey, A.J., Ventricular size in patients with presenile dementia of the Alzheimer's type, Arch. Neurol., 41 (1984) 1258-1263. 2 Brinkman, S.D., Sarwar, M., Levin, H.S. and Morris, H.H., Quantitative index of computed tomography in dementia or normal aging, Neuroradiology, 138 (1981) 89-92. 3 de Leon, M.J., Ferris, S.H., George, A.E., Reisberg, B., Kricheff I.I. and Gershon, S., Computed tomography evaluation of brain-behaviour relationships in senile dementia of Alzheimer's type, Neurobiol. Aging, 1 (1980) 69-79. 4 de Leon, M.J. and George, A.E., Computed tomography in aging and senile dementia of the Alzheimer type, Adv. Neurol., 38 (1983) 103-122. 5 Fischer, H.W., Rinck, P.A., Van Haverbeke, Y. and Muller, R.N., Nuclear relaxation of human brain gray and white matter: analysis of field dependence and implications for MRI, Magn. Resort. Med., 16 (1990) 317-334. 6 Foster, M.A. and Hutchison J.M.S. (Eds.), Practical NMR Imaging, IRL press, Oxford, 1987. 7 Fukushima, M., Histometric and histochemical studies of the hypothalamo-hypophyseal system of spontaneously hypertensive rats and rats with experimental hypertension, Jpn. Circ. J., 33 (1968) 485-516. 8 Gado, M., Patel, J., Hughes, C.P., Danziger, W. and Berg, L., Brain atrophy in dementia judged by CT Scan ranking, Am. J. Neuroradiol., 4 (1983) 449-500. 9 Johnstone, E.C., Frith, C.D., Crow, T.J., Husband, J. and Kreel, L., Cerebral ventricular size and cognitive impairment in chronic schizophrenia, Lancet, ii (1976) 924-925. 10 Lang, R.E., Rascher W., Unger, T. and Ganten, D., Reduced content of vasopressin in the brain of spontaneously hypertensive rats as compared to norrnotensive rats, Neurosci. Lett., 23 (1981) 199-201. 11 Last, R.J. and Tompsen, D.H., Casts and cerebral ventricles,
3-7.5 months. No correlation between blood pressure values and ventricular vol. was found by comparing rats of the same age groups. Thus, our results support Ritter et al.'s suggestion 19 that hypertension and ventricular enlargement seem to result from inbreeding of dissociated genetic trait. Several h u m a n neural degenerative disorders are associated with ventricular dilation 5'8'9'12. Pathological studies of normal aging brain have revealed ventricular enlargement with concurrent loss of brain weight and vol. 4'11'21. H o w e v e r , in A l z h e i m e r disease the enlargement of the ventricular size was found to be superimposed upon the age related changes 3'4. In some additional degenerative disorders it was shown that the elevation in the ventricular size is associated with the magnitude of the cognitive deficit 1'2'8'el. Thus, it is possible that the elevation in ventricular vol. of the S H R brain is a consequence of (or predisposal to) a brain degeneration process, which may contribute to the already reported abnormalities of SHR's 13~17.
Acknowledgements. We gratefully acknowledge helpful suggestions from Prof. Menachem Segal. This work was supported by the Israel Institute for Psychobiology, Charles E. Smith Family Foundation.
Br. J. Surg., 40 (1953) 525-543. 12 Massman P.J., Bigler E.D., Cullum, S.M. and Naugle, R.I., The relationship between cortical atropy and ventricular volume, Int. J. Neurosci., 30 (1986) 87-99. 13 McCarty, R. and Kopin, I.J., Patterns of behavioral development in spontaneously hypertensive rats and Wystar Kyoto normotensive controls, Dev. Psychobiol., 12 (1979) 239-243. 14 Morris, M., Wren, J.A. and Sundberg, D.K., Central neural peptides and catecholamines in spontaneous and DOCA/salt hypertension, Peptides, 2 (1981) 207-211. 15 Nelson, D.O. and Boulant, J.A., Altered CNS neuroanatomical organization of spontaneously hypertensive (SHR) rats, Brain Research, 226 (1981) 119-130. 16 Norman, A.B., Thomas, S.R., Pratt, R.G., Samaratunga, R.C. and Sanberg, P.R., Magnetic resonance imaging of rat brain following kainic acid-induced lesions and fetal strial tissue transplants, Brain Research, 483 (1989) 188-191. 17 Okamoto, K., Spontaneous hypertension in rats, Int. Rev. Exp. Pathol., 7 (1969) 675-681. 18 Okamoto, K. and Aoki, K., Development of a strain of spontaneous hypertensive rats, Jpn. Orc. J., 27 (1963) 282-293. 19 Ritter, S., Dinh, T.T., Stone, S. and Ross, N., Cerebroventricular dilation in spontaneously hypertensive rats (SHRs) is not attenuated by reduction of blood pressure, Brain Research, 450 (1988) 354-359. 20 Ritter, S. and Dinh, T.T., Progressive postnatal dilation of brain ventricles in spontaneously hypertensive rats, Brain Research, 370 (1986) 327-332. 21 Tomlinson, B.E., The structural and quantitative aspects of the dementias. In P.J. Roberts (Ed.), Biochemistry of Dementia, Wiley, New York, 1980, pp. 15-52. 22 Williams, S.C.R., Harries, N.G. and Jones, H.C., Magnetic resonance imaging of congenital hydrocephalus in the developing rat brain, Abstracts (WIP), Society of Magnetic Resonance in Medicine, Ninth Annual Scientific Meeting, New York. 1990, 1170 pp.
Brain Research, 574 (1992) 224-228 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$05.1X)
BRES 17509
Quantitation of ventricular size in normal and spontaneously hypertensive rats by magnetic resonance imaging Peter Bendel I and Raya Eilam 2 Departments of 1Chemical Physics and 2Neurobiology, The Weizmann Institute of Science, Rehovot (Israel) (Accepted 22 October 1991) Key words: Magnetic resonance imaging; Brain ventricular dilation; Spontaneously hypertensive rat; Wystar-Kyoto rat
Magnetic resonance imaging (MRI) performed at high field (4.7 Tesla), and high spatial resolution (0.6 mm slice thickness, 0.18 mm inplane) enabled noninvasive quantitative measurement of the ventricular vol. in live rats. Comparing the results for 15 male Wystar-Kyoto (WKY) rats, aged 2.5-10 months, with those from 17 spontaneously hypertensive rats (SHR), clearly confirmed the previously reported elevated ventricular vols. in the SHR strain. A significant difference in ventrieular vol. between the two strains was detected above the age of 3 months. For mature animals above the age of 6 months the mean vol. in the SHR strain was elevated by about a factor of two compared to the WKY control animals. INTRODUCTION The spontaneously hypertensive rat (SHR) strain has been used extensively as an animal model for essential hypertension in the human. Since the S H R was derived by selective inbreeding from the W K Y strain 18, W K Y are usually considered to be the appropriate normotensive controls for the SHR. These hypertensive rats are characterized by a broad spectrum of anatomical, histological, as well as neuroendocrine deviations 7'1°'14. It was recently reported that brain ventricular size of the S H R is significantly elevated in comparison to ageand sex-matched W K Y rats. The ventricular dilation is associated with a loss of brain tissue 2° and reduced brain weight and vol. 19, which may underlie some of the behavioral abnormalities of SHR. It was found that ventricular enlargement develops abruptly between 4 and 8 weeks of age, thereafter progressing at a slower rate. This pathology was found to occur in the absence of visible obstruction within the ventricular space, and does not appear to result from permeability changes of the blood-brain barrier. The conventional methodology for assessing ventricular size, which led to the findings mentioned above 19'2° involved sacrificing the animal, followed by perfusion and cutting of brain sections. Magnetic resonance imaging (MRI) has the potential for completely noninvasive examination of the brain anatomy. It therefore enables us to follow the development of anatomical pathology in the same animal, which could be correlated with con-
comittant physiological and behavioral alterations. The aim of our study was to develop M R I methods which would permit quantitative noninvasive measurement of brain ventricular vols. and validate those methods. M R I is a powerful radiological diagnostic method for humans, particularly in the central nervous system 6. The spatial resolution typically achieved in medical imaging (slice thickness of the order of 5 ram, in-plane resolution of the order of 1 mm) is not suitable for detecting the fine anatomical details, in particular the ventricular structure, within the rat brain. U p to now, M R images of rat brain 16, or attempts to quantitate enlarged rat brain ventricles by MR122 did not achieve the resolution necessary for measuring ventriclular size in animals with normal or moderately enlarged ventricles. We succeeded in attaining improved resolution, enabling clear demonstration of the ventricles. The key factors which permitted this were increased signal-to-noise ratio (S/N) achieved due to the high magnetic field, radio frequency (rf) and gradient hardware which is matched to the size of small samples, and by applying imaging protocols which maximize the contrast between cerebro-spinal fluid (CSF) and surrounding brain tissue.
MATERIALS AND METHODS The MR images presented in this report were obtained on a Bruker Biospec 47/30 NMR system, operating at 4.7 Tesla field strength (resonance frequency for protons: 200 MHz) with a 30 cm diameter horizontal access bore. The S/N was maximized by using a home-built rf coil of 5 cm diameter which closely fits over the
Correspondence: P. Bendel, Department of Chemical Physics, The Weizmann Institute of Science, Rehovot 76100, Israel.
225 Fig. 1. Example of a sagittal localization image, obtained in 1.3 minutes with repetition time (TR) of 600 ms and echo time (TE) of 17 ms. The slice thickness is 2 mm. The coronal slices for the study which will be used for the ventricle size measurements are selected on this image by interactive software.
head of the animal. After initial evaluation with different imaging protocols, using varying parameters, we settled on the following routine procedure for obtaining the MR images: The rats were anesthetized with pentobarbital (about 40 mg/kg i.p.). A rapid orientation image was obtained (in about 2 min of scan time) in a sagittal plane, and the coronal slices for the subsequent study were selected on this orientation image, perpendicular to the anteriorposterior axis of the brain (Fig. 1). Twelve to 18 slices were selected to cover the region between the septal caudate-putamen nuclei and the substantia nigra nucleus. Slice thickness was 0.6 mm in most studies, with an interslice gap of 0.09 mm. (Some of the studies were conducted with slice thickness 0.87 mm and gaps of 0.13
~i
•
i ~¸
~
Fig. 2. Consecutive representative sfices from T2-weighted magnetic resonance imaging (MRI) scans (TR = 3.3 s, TE = 68 ms) for a spontaneously hypertensive rat (SHR) (right) and WKY (left) rat. The order of advance from posterior to anterior is left to right and top to bottom. The small square for each slice represents an area of 2.25 × 2.25 cm 2. Slice thickness is 0.6 ram, and the center-to-center distance between slices is 0.69 ram. The WKY rat is listed #10, and the SHR #13 in Table I.
226 80
v
60
~,
4o
~
20
IT/
~t
> 0
2.5
3-5 Age (months)
lated by the computer and displayed. Total ventricular vols. were determined by adding the areas of all the ventricular structures on the monitored slices, multiplied by the distance between adjacent slice centers. This post-processing procedure took an additional half hour per study. RESULTS T h e results for c o m p u t e d v e n t r i c u l a r vols. f r o m a total of 27 a n i m a l studies are t a b u l a t e d in Table I and graphically displayed in Fig. 3. T h e values w e r e c o m p u t e d as i n d i c a t e d in the p r e c e d i n g M a t e r i a l s and M e t h -
6-8
ods section. In 10 of t h e studies, the v e n t r i c u l a r vol. was
Fig. 3. Mean ventricular vol. vs age, summarizing the results in Table I, for WKY (empty boxes) and SHR (shaded boxes) rats. Significant differences between groups are shown by asterixes.** P < 0.05.
c o m p u t e d twice by the s a m e o b s e r v e r (P.B.) to assess i n t r a r a t e r reliability (r = 0.989). T h e s a m e 10 studies w e r e also a n a l y z e d by a s e c o n d o b s e r v e r ( R . E . ) , yielding r = 0.979 for the c o r r e l a t i o n b e t w e e n the values of
mm.) The field-of-view (FOV) was 4.5 cm on a 256 x 256 matrix. Images were acquired with a standard 2DFT spin-echo protocol, using a repetition time (TR) of 3.3 s, a single echo with echo time (TE) of 68 ms, and 2 averages per cycle. The total imaging time for the coronal images was therefore 29 rain and the total experimental study time for each animal (including positioning, calibrations and orientation image) was about 45 rain. The images acquired under the above conditions are T2-weighted, with the CSF, which has a relatively long T2 relaxation time, appearing at much higher intensity than the surrounding brain tissue, thus clearly delineating the outline of the ventricular structure. Fig. 2 shows the results of such T2-weighted coronal slices, displaying the images from representative studies for an SHR and a WKY rat. The standard image-display software on the instrument allows the interactive delineation of arbitrary regions-of-interest (ROI) on the images (by traekbaU). The areas of these ROl's were then calcu-
o b s e r v e r II and the m e a n values of o b s e r v e r I. Considering the scatter in all t h e s e m e a s u r e m e n t s (and including several i m a g i n g e x p e r i m e n t s r e p e a t e d on the s a m e a n i m a l on the s a m e day), a r o o t - m e a n - s q u a r e e r r o r of _+ 5 m m 3 for the r e p o r t e d v e n t r i c u l a r vols. was calculated. T h e m o s t significant p o t e n t i a l e r r o r source for the absolute accuracy of t h e results is the finite c e n t e r - t o - c e n ter distance (of 0.69 m m ) b e t w e e n a d j a c e n t slices. If the v e n t r i c u l a r structure c h a n g e s significantly o v e r this scale, only an a v e r a g e structure will p r o j e c t o n t o the two-dim e n s i o n a l i m a g e in e a c h slice (the so-called 'partial volu m e ' effect), which m a y cause errors in the c o m p u t e d
TABLE I Experimental results for ventricular volumes WKY #
Age (months)
Volume a (mm 3)
Blood pressure
SHR #
Age (months)
Volume ~ (ram 3)
Blood pressure
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
2.5 2.5 2.5 3 3 3 4 4 5 5 6 7 8 8 10
19 13 14 26 29 25 36 26 24 31 29 35 36 36 30
110 115 115 110 130 130
1 2 3 4 5 6 7 8 9 9a 10 11 12 13 14 14a 15 16 17 18 19
2 2 2.5 2.5 2.5 3 3 5 5 6 6 6 6 6 6 7 7.5 7.5 7.5 7.5 13
8 8 16 16 14 38 39 37 36 52 68 68 58 85 54 61 60 51 60 26 56
120 115 130 155 155 185 165 200 200 200 200 205
120 120 135
a Experimental uncertainty: + 5 mm 3. SHR # 9a and 14a are follow-up studies of the same animals as in 9 and 14.
160 220 240 240 235 185
227
Fig. 6. Comparison between a spin density weighted (left) and T2weighted image (right). The images are from a double-echo scan, with TR = 3.3 s, and echo times as indicated.
Fig. 4. Comparison between 0.6 mm thick MRI slices from the live animal (left) and brain sections (thickness: 0.1 mm) obtained sacrificing the same animal (SHR #13), followed by perfusion. Bar = 0.5 cm.
vol. It is difficult to quantitate this possible error in accuracy, but clearly it should be become worse with increasing slice thickness and interslice gap. We therefore conducted some experiments at an additional different setting of 0.87 m m slice thickness and 0.13 m m gap, but found no significant change in the measured ventricular vols., indicating that if a systematic error exists, it should be within the limits of the exPerimental uncertainty. In mature W K Y rats there was no significant variation of ventricular vol. with age. We found the mean vols. to be 28.1 + 1.5 m m 3 (n = 7) for animals aged 3-5 months, and 33.2 + 3.8 mm 3 for 6 - 8 months old rats. For the monitored S H R rats an increase in ventricular vol. was clearly discernible, up to the age of about 7 months. The mean values were 37.5 + 6.6 m m 3 for rats of age 3-5
Fig. 5. Comparison between a /'1- (left) and T2-weighted scan (right). Slice thickness is 0.6 mm, and TRZI'E as indicated. The square for each image is 2.25 x 2.25 cm2. Notice the faint dark appearance of the ventricles on the left image, vs the strong positive contrast between the ventricles and the brain on the right. The bright appearance of the ventricles is due to the long T2 values of the CSE
months (n = 4), and 58.4 + 4.3 m m 3 for ages 6 - 7 months (n = 11). This trend was also reflected in the resuits obtained from two animals in which follow-up studies were conducted a month after the original study. However, the results for two individual studies (26 m m 3 and 85 m m 3) diversed widely from the rest in their age range, indicating a possible strong individual variation of ventricular size for S H R rats in the same age group. A significant difference between the hypertensive and normotensive groups (see Fig. 3) was found in the adult groups (above 3 months), but not between the young animals. Perfused brain sections of W K Y rats and SHR's were compared with the corresponding M R I slices to qualitatively demonstrate the agreement between the M R images and the anatomy as reflected on the perfused sections (Fig. 4). DISCUSSION The crucial factors which enabled the use of M R to image the ventricles in rat brains were maximal S/N, as well as imaging protocol parameters that have fine enough resolution (in terms of slice thickness and FOV) to visualize the details of the ventricular structure and create adequate contrast between CSF and surrounding brain tissue. The S/N necessary to produce 0.6 m m thick slices and 0.18 m m 2 in-plane pixel size could only be achieved at field strengths which are considered high for medical imaging (4.7 Tesla in our case, whereas wholebody medical imagers reach 2 Tesla at most). Close-fitting coils, which maximize the filling factor for the brain, are also important. Finally, the optimal parameters (TR, TE) for the imaging protocol should be chosen with care. A t lower magnetic fields, such as those used for medical imaging, CSF is usually visualized easily by its low signal intensity on Tl-weighted scans. However, at higher fields the 7"1 values for brain gray and white mat-
228 ter increase considerably 5, much m o r e than for C S E A s a result, the contrast on Tl-weighted brain images at high fields is very p o o r , as d e m o n s t r a t e d in Fig. 5, which compares T 1- and T2- weighted scans of the same slice. On spin-density weighted scans, the C S F to brain contrast is also not as g o o d as in T2-weighted images. This is shown in Fig. 6, which displays the first and second echo from a double echo sequence. One of the well known p h e n o m e n a that characterizes S H R ' s is a gradual d e v e l o p m e n t of hypertension as a function of age 17. During the m a t u r a t i o n process, significant loss of brain weight and vol., as well as reduction in cell cross-section areas, c o m p a r e d with normotensive ( W K Y ) rats has also been r e p o r t e d 7'15. Interestingly, S H R were found to display brain ventricular enlargement during the same p e r i o d 2°. This pathology could not have been the direct result of peripheral elevation of blood pressure, since chronic reduction of blood pressure in S H R by capnopnil t r e a t m e n t failed to attenuate the dilation. By the same token, experimentally produced hypertension did not lead to increase in brain ventricular size in W K Y rats 19. In the present study we also found a d e v e l o p m e n t of the cerebral ventricular enlargement in S H R rats aged
REFERENCES 1 Albert, M., Naeser, M.A., Levine, H.L. and Garvey, A.J., Ventricular size in patients with presenile dementia of the Alzheimer's type, Arch. Neurol., 41 (1984) 1258-1263. 2 Brinkman, S.D., Sarwar, M., Levin, H.S. and Morris, H.H., Quantitative index of computed tomography in dementia or normal aging, Neuroradiology, 138 (1981) 89-92. 3 de Leon, M.J., Ferris, S.H., George, A.E., Reisberg, B., Kricheff I.I. and Gershon, S., Computed tomography evaluation of brain-behaviour relationships in senile dementia of Alzheimer's type, Neurobiol. Aging, 1 (1980) 69-79. 4 de Leon, M.J. and George, A.E., Computed tomography in aging and senile dementia of the Alzheimer type, Adv. Neurol., 38 (1983) 103-122. 5 Fischer, H.W., Rinck, P.A., Van Haverbeke, Y. and Muller, R.N., Nuclear relaxation of human brain gray and white matter: analysis of field dependence and implications for MRI, Magn. Resort. Med., 16 (1990) 317-334. 6 Foster, M.A. and Hutchison J.M.S. (Eds.), Practical NMR Imaging, IRL press, Oxford, 1987. 7 Fukushima, M., Histometric and histochemical studies of the hypothalamo-hypophyseal system of spontaneously hypertensive rats and rats with experimental hypertension, Jpn. Circ. J., 33 (1968) 485-516. 8 Gado, M., Patel, J., Hughes, C.P., Danziger, W. and Berg, L., Brain atrophy in dementia judged by CT Scan ranking, Am. J. Neuroradiol., 4 (1983) 449-500. 9 Johnstone, E.C., Frith, C.D., Crow, T.J., Husband, J. and Kreel, L., Cerebral ventricular size and cognitive impairment in chronic schizophrenia, Lancet, ii (1976) 924-925. 10 Lang, R.E., Rascher W., Unger, T. and Ganten, D., Reduced content of vasopressin in the brain of spontaneously hypertensive rats as compared to norrnotensive rats, Neurosci. Lett., 23 (1981) 199-201. 11 Last, R.J. and Tompsen, D.H., Casts and cerebral ventricles,
3-7.5 months. No correlation between blood pressure values and ventricular vol. was found by comparing rats of the same age groups. Thus, our results support Ritter et al.'s suggestion 19 that hypertension and ventricular enlargement seem to result from inbreeding of dissociated genetic trait. Several h u m a n neural degenerative disorders are associated with ventricular dilation 5'8'9'12. Pathological studies of normal aging brain have revealed ventricular enlargement with concurrent loss of brain weight and vol. 4'11'21. H o w e v e r , in A l z h e i m e r disease the enlargement of the ventricular size was found to be superimposed upon the age related changes 3'4. In some additional degenerative disorders it was shown that the elevation in the ventricular size is associated with the magnitude of the cognitive deficit 1'2'8'el. Thus, it is possible that the elevation in ventricular vol. of the S H R brain is a consequence of (or predisposal to) a brain degeneration process, which may contribute to the already reported abnormalities of SHR's 13~17.
Acknowledgements. We gratefully acknowledge helpful suggestions from Prof. Menachem Segal. This work was supported by the Israel Institute for Psychobiology, Charles E. Smith Family Foundation.
Br. J. Surg., 40 (1953) 525-543. 12 Massman P.J., Bigler E.D., Cullum, S.M. and Naugle, R.I., The relationship between cortical atropy and ventricular volume, Int. J. Neurosci., 30 (1986) 87-99. 13 McCarty, R. and Kopin, I.J., Patterns of behavioral development in spontaneously hypertensive rats and Wystar Kyoto normotensive controls, Dev. Psychobiol., 12 (1979) 239-243. 14 Morris, M., Wren, J.A. and Sundberg, D.K., Central neural peptides and catecholamines in spontaneous and DOCA/salt hypertension, Peptides, 2 (1981) 207-211. 15 Nelson, D.O. and Boulant, J.A., Altered CNS neuroanatomical organization of spontaneously hypertensive (SHR) rats, Brain Research, 226 (1981) 119-130. 16 Norman, A.B., Thomas, S.R., Pratt, R.G., Samaratunga, R.C. and Sanberg, P.R., Magnetic resonance imaging of rat brain following kainic acid-induced lesions and fetal strial tissue transplants, Brain Research, 483 (1989) 188-191. 17 Okamoto, K., Spontaneous hypertension in rats, Int. Rev. Exp. Pathol., 7 (1969) 675-681. 18 Okamoto, K. and Aoki, K., Development of a strain of spontaneous hypertensive rats, Jpn. Orc. J., 27 (1963) 282-293. 19 Ritter, S., Dinh, T.T., Stone, S. and Ross, N., Cerebroventricular dilation in spontaneously hypertensive rats (SHRs) is not attenuated by reduction of blood pressure, Brain Research, 450 (1988) 354-359. 20 Ritter, S. and Dinh, T.T., Progressive postnatal dilation of brain ventricles in spontaneously hypertensive rats, Brain Research, 370 (1986) 327-332. 21 Tomlinson, B.E., The structural and quantitative aspects of the dementias. In P.J. Roberts (Ed.), Biochemistry of Dementia, Wiley, New York, 1980, pp. 15-52. 22 Williams, S.C.R., Harries, N.G. and Jones, H.C., Magnetic resonance imaging of congenital hydrocephalus in the developing rat brain, Abstracts (WIP), Society of Magnetic Resonance in Medicine, Ninth Annual Scientific Meeting, New York. 1990, 1170 pp.
