- Email: [email protected]
Functional 2D and 3D magnetic resonance imaging of motor cortex stimulation at high spatial resolution using standard 1.5 T imager
Magnelic Resonance Imaging. Vol. 12, pp. 9-15. 1994 Printed in the USA. All rights reserved.
Copyright 0
0730-725X/94 $6.00 + .W 1993 Pergamon Press Ltd.
0 Original Contribution
FUNCTIONAL 2D AND 3D MAGNETIC RESONANCE IMAGING OF MOTOR CORTEX STIMULATION AT HIGH SPATIAL RESOLUTION USING STANDARD 1.5 T IMAGER LOTHAR R. SCHAD,* FREDERIK WENZ,* MICHAEL V. KNOPP,* KLAUS BAUDENDISTEL,* EDGAR M~~LLER,~ AND WALTER J. LORENZ* *Department of Radiology, German Cancer Research Center, Heidelberg, Germany and tSiemens Medical Division, Erlangen, Germany This paper reports the effects of motor cortex stimulation of normal volunteers using conventional MR imaging techniques on standard 1.5 T clinical scanner. Improvement in signal-to-noise (S/N) ratio has been achieved by using a commercially available eye/ear surface coil with a loop of 8.5 cm in diameter. Magnet shimming with all first order coils was performed to the volunteer’s head resulting in a magnetic field homogeneity of about 0.1-0.2 ppm. The imaging technique used was an optimized conventional 2D and 3D, first order flow rephased, gradient-echo sequence (FLASH) with fat-suppression and reduced bandwidth (16-28 Hz/pixel) and TR = 80120 ms, TE = 60 ms, flip angle = 40°, matrix = 128 x 128, FOV = 150-250 mm, slice-thickness = 2-5 mm, NEX = 1, and a total single scan time for one image of about 12-16 s. In the 3D FLASH measurements, a slab of 32 mm thickness with 16 partitions was evaluated. The motor cortex stimulation was achieved by touching each finger to thumb in a sequential, self-paced, and repetitive manner. During stimulation, an increase in signal of order lo-20% was detected in the motor and sensory cortex due to reduced partial volume effects and optimized S/N for the measurements at small voxel size. 3D FLASH imaging at high spatial resolution shows good anatomical correlation of signal increase with gray matter of the motor and sensory cortex. The reported data demonstrate the technical feasibility of functional 2D and 3D MR imaging at high spatial resolution using optimized conventional sequences and equipment.
Keywords: Magnetic resonance imaging, techniques; Functional imaging; Motor cortex stimulation.
INTRODUCTION
effect), while oxygenized haemoglobin is not paramagnetic. On the other hand, regional cerebral blood volume (rCBV) and regional cerebral blood flow (rCBF) increase during stimulation which have been demonstrated by PET measurements showing an increase in rCBF during stimulation of about 5-20% .4 The net effect of regional MR signal change (TT effect modulated by chances in rCBV and rCBF) can be measured by sequences that specifically enhance this Tz effect by using long echo times. At the Eleventh Annual Scientific Meeting of the Society of Magnetic Resonance in Medicine (1992) in Berlin, recently reported studies were presented on completely noninvasive MR imaging of human brain activity based on either blood flow or blood oxygenation changes.5-‘4 Most of the investigators were using echo planar imaging (EPI) methods requiring special hardware equipment. Stimulation
The visualization of brain function and its reaction to external stimuli has always been of great interest both to clinical as well as scientific research. Until recently, this area was the domain of positron emission tomography (PET). New measuring techniques have demonstrated that functional imaging of brain perfusion is also possible with magnetic resonance (MR) imaging.lm3The goal in these studies was to employ the inherent sensitivity of MR to local changes in magnetic susceptibility which are induced by paramagnetic particles like Gadolinium-DTPA, the well known contrast agent in MR. Another even more noninvasive approach is to use the changes and paramagnetic effects of haemoglobin. Deoxygenized haemoglobin is paramagnetic and slightly distorts the magnetic field around it (r,* RECEIVED 2/10/93; ACCEPTED 7/6/93. Address correspondence to PD Dr. L.R.
Schad,
schungsschwerpunkt zentrum, Postfach
For9
Radiologie, Deutsches Krebsforschungs101949, D-69009 Heidelberg, Germany.
10
Magnetic Resonance Imaging 0 Volume 12, Number
measurements on standard scanners using conventional imaging techniques were reported rarely and are essentially preliminary.‘5-‘8 The present study was undertaken to evaluate the feasibility of functional 2D and 3D MR imaging of the human brain at high spatial resolution and with improved signal-to-noise (S/N) ratio on the basis of optimized conventional FLASH (Fast Low Angle Shot) sequences” using standard 1.5 T whole body equipment. MATERIALS
AND METHODS
Instrumentation MR imaging was performed on a 64 MHz Magnetom (Siemens, Erlangen, Germany) superconducting whole body imager. Improvement in S/N of the stimulation measurements has been achieved by using a commercially available eye/ear surface coil with a loop of 8.5 cm in diameter. High loaded Q-values results in an improved S/N ratio and thus high-quality images. Magnet shimming with all first order coils was performed to the volunteer’s head resulting in a magnetic field homogeneity of about 0.1-0.2 ppm. The imaging technique used was an optimized conventional 2D and 3D, first order flow rephased, gradient-echo sequence (FLASH) with fat-suppression and reduced bandwidth (16-28 Hz/pixel) for getting optimal S/N ratio. The sequence parameters used are: TR = 80-120 ms, TE = 60 ms, flip angle = 40”, matrix = 128 x 128, FOV = 150-250 mm, slice-thickness = 2-5 mm, NEX = 1, and a total single scan time for one image of about 12- 16 s. In the 3D FLASH measurements, a slab of 32 mm thickness with 16 partitions was evaluated resulting in a total measuring time for the 3D experiment of about 4 min. All FLASH experiments were anatomically correlated with conventional T1-weighted spin-echo (SE) images. The MR computer (MICRO-VAX, DEC, Maynard, Massachusetts, U.S.A) was connected directly (DEC-NET, DEC, link) to a central computer (VAXWORKSTATION/3600, DEC) where the software program developed by our group for 2D and 3D functional mapping was implemented. After data acquisition, typically 60-100 images were transferred to the central computer. Clinical Study At the time of publication, 20 volunteers were investigated. For orientation, sagittal FLASH images were measured firstly. Based on these sagittal images, 7’,-weighted SE imaging (TR/TE 600/15) has been performed in an oblique plane in between axial and coronal planes with a tilting angle of about 10”. The slices
1, 1994
with the best visualization of the primary motor cortex were chosen for dynamic 2D imaging and high resolution 3D mapping of the stimulation experiments. The motor cortex stimulation was achieved by touching each finger to thumb in a sequential, self-paced, and repetitive manner. RESULTS Figure 1A shows an oblique (axial to coronal: loo) SE image of a normal volunteer with high spatial resolution (FOV = 150 mm, slice-thickness = 2 mm, matrix = 256 x 256) using the surface coil. Figure 1B demonstrates the corresponding result of a dynamic 2D FLASH stimulation experiment (FOV = 150 mm, slicethickness = 2 mm, matrix = 128 x 128) showing a difference image between averages of all stimulation (30 images) and nonstimulation (30 images) images. The stimulation experiment starts with the measurement of 10 control images (without finger movement) followed by 10 corresponding stimulation images (with finger movement of the right hand). This stimulus alternation was repeated for three cycles. Comparing the difference image with the T, -weighted SE image a signal increase produced by this stimulation (touching each finger to thumb) occurred in areas predominantly occupied by the grey matter of primary left motor and sensory cortex. Figure 2A reflects two representative signal-time curves in the primary motor cortex (region of interest mean values including about 20 pixels) and the sensory cortex during the stimulation experiments. The data show a significant increase in signal of order lo-20%. The effects in signal change during stimulation can be seen in one pixel only with an increase in signal of order 20% (Fig. 2B). Figure 2C represents the results of the same stimulation experiment (signal-time curve in the primary motor cortex), but of another volunteer where a clear linear signal increase during stimulation was observed. Figure 3 shows eight central slices out of a total of 16 slices as a typical example of a 3D FLASH stimulation experiment of a normal volunteer using the surface coil. Thus, oblique (axial to coronal: 10’) SE images (FOV = 200 mm, slice-thickness = 2 mm, matrix = 256 x 256) were overlayed by the corresponding difference images (green color) of the 3D FLASH stimulation experiment (FOV = 200 mm, slab-thickness = 32 mm/16 partitions, matrix = 128 x 128). In particular, the 3D stimulation experiment starts with a 3D FLASH measurement (16 slices in 4 min) without finger movement, followed by a corresponding 3D FLASH measurement (16 slices in 4 min) with finger movement of the right hand. This stimulus alternation was repeated for eight cycles. The difference images (Fig. 3,
Functional 2D and 3D MRI at high spatial resolution 0 L.R.
(4
SCHAD
ET AL.
11
@I
Fig. 1. 2D stimulation experiment. (A) Oblique (axial to coronal: 10”) SE image (TRITE 600/15) of a normal volunteer with high spatial resolution (FOV = 150 mm, TH = 2 mm, matrix = 256 x 256) using the surface coil. (B) Difference image (FOV = 150 mm, TH = 2 mm, matrix = 128 x 128) between averages of all stimulation (30 images) and nonstimulation (30 images) images showing signal increase produced by the stimulation (touching each finger to thumb) occurred in areas predominantly occupied by the grey matter of primary left motor and sensory cortex.
green overlay color) were calculated by averaging of all 3D stimulation images (8 images at every slice position) and nonstimulation images, and subtraction of each other. The 3D FLASH stimulation measurement at high spatial resolution demonstrates again that signal increase produced by this kind of stimulation (touching each finger to thumb) occurred in areas predominantly occupied by the grey matter of primary left motor and sensory cortex. DISCUSSION As expected, stimulation experiments at high spatial resolution show a clear enhancement of the observed stimulation effect (about 20%) and an improvement in detectability. As seen from the head coil experiments at larger voxel size,” an increase in signal intensity due to larger voxels does not lead directly to improvements in the detectability of signal change because of partial volume averaging with nonstimulated areas of the brain. Of course, reduced voxel size lead to a reduction of S/N. Improvement in S/N can be achieved by using surface coils with high loaded Q-values and by optimization of the measuring sequences. Reduction in the bandwidth to about 20 Hz/pixel does significantly increase the
S/N. On the other hand, chemical shift artefacts caused by the fat mainly from the outer side of the head lead to a shift between fat and water images of about 10 pixels in read-out direction. This chemical shift artefact can shift parts of the fat image onto areas of motor or sensory cortex and thus obscure the stimulation effects. To overcome this problem we used standard fat suppression in the FLASH sequences. In our experience, an effective fat suppression can be reached for the head coil and surface coil experiments at a magnetic field homogeneity of about 0.1-0.2 ppm. 2D and 3D FLASH stimulation experiments at high spatial resolution compared with Tl-weighted SE images show a clear anatomical correlation between signal increase and grey matter of primary motor cortex and sensory cortex. Although the 3D FLASH experiments (Fig. 3) show activation along each sulcus, the major effects are confined to the outer layers of grey matter along motor and sensory cortex. The additional highlighted regions may arise from pulsation of cerebrospinal fluid or may even be caused by flow effects and vascular pulsation artefacts of large vessels, that is, superior sagittal sinus. However, yet limited in the number of volunteers, the available data reveal a remarkable interindividual reproducibility of the expected
Magnetic Resonance Imaging 0 Volume 12, Number 1, 1994
12
600
800 time
[set]
surface coil 1 DiXd
25C
200
600
400
200
0
800 time
[set]
W 130’-
7
.
surface coil 20pki 120.
ti B # #.
ilO0
rl, 0
#
l.,
.o -0
l 0
stirmalinl
800
600
time
Fig. 2. 2D stimulation
[eec]
dynamic. (A) Typical signal-time curves of the primary motor cortex and sensory cortex (region of interest mean values including about 20 pixels) showing a significant increase in’signal of order lo-20% during stimulation. (B) Effects in signal change can be seen in one pixel only with an increase in signal of about 20% during stimulation. (C) Results of the same stimulation experiment (signal-time curve in the primary motor cortex) but of another volunteer where a clear linear signal increase during stimulation was observed.
0)
09
:E)
Fig. 3. 3D stimulation experiment. (A-H) Eight central slices out of a total of 16 slices of a typical 3D FLASH stimulation experiment of a normal volunteer using the surface coil (FOV = 200 mm, SLAB = 32 mm/16 PART, matrix = 128 x 128). Oblique (axial to coronal: 10”) SE images were overlayed by the corresponding difference images (green color) of the 3D FLASH stimulation experiment showing again a clear anatomical correlation between signal increase produced by this kind of stimulation (touching each finger to thumb) and areas predominantly occupied by the grey matter of primary left motor and sensory cortex. A maximum intensity projection (MIP) of the stimulated areas seems to be possible.
(4
W
14
Magnetic Resonance Imaging 0 Volume 12, Number 1, 1994
activation maps. The observed increase in signal intensity in the activated areas of the brain is potentially explained qualitatively by findings previously confirmed by PET observations,4 that during local brain stimulation, oxygen delivery to the activated region exceeds metabolic need. The overabundance of oxygen-rich blood leads to a decrease in the local oxygen extraction fraction, a local increase in O2 utilization, and a local decrease in the concentration of deoxyhemoglobin. Thereby deoxyhemoglobin acts as an effective endogenous contrast agent where signal variation during stimulation is caused by differences in the magnetic susceptibility of blood induced by the presence of paramagnetic deoxyhemoglobin in red cells. A decrease in deoxyhemoglobin concentration decreases the vesseltissue susceptibility differential, allowing increased spin coherence and thus increased signal in gradient-echo imaging (increase of T,*). Our data (20 volunteers) show different signal behaviour during stimulation: most of the volunteers showed a more or less constant enhanced signal during stimulation, but in some cases a clear linear signal increase was observed (Fig. 2C). At the moment, the neurophysiological explanation of this different signal behaviour is unclear and needs some further clinical investigations completed by PET experiments. In conclusion, our data demonstrate the technical feasibility of 2D and 3D functional MR imaging at high spatial resolution on standard clinical 1.5 T scanner using optimized gradient-echo sequences and coil equipment. 2D time course gradient-echo imaging is a powerful new noninvasive tool for assessment of regional cerebral activation with high temporal and spatial resolution. 3D functional imaging is another powerful tool to image the stimulation effects of the whole brain cortex and a maximum intensity projection (MIP), the well known projection method in MR angiography, seems to be possible.
5.
6.
7.
8.
9.
10.
11.
REFERENCES Ogawa, S.; Lee, T.M. ; Nayak, A.S.; Glynn, P. Oxygenation-sensitive contrast in magnetic resonance image of rodent brain at high magnetic fields. Magn. Reson. Med. 14:68-78; 1990. Rosen, B.R.; Belliveau, J.W.; Vevea, J.M.; Brady, T.J. Perfusion imaging with NMR contrast agents. Magn. Reson. Med. 14:249-265; 1990. Belliveau, J.W.; Kennedy, D.N.; McKinstry, R.C.; Buchbinder, B.R.; Weisskoff, R.M.; Cohen, M.S.; Vevea, J.M.; Brady, T.J.; Rosen, B.R. Functional mapping of the human visual cortex by magnetic resonance imaging. Science 254:716-719; 1991. Fox, P.T.; Raichle, M.E. Focal physiological uncoupling
12.
13.
of cerebral blood flow and oxidative metabolism during somatosensory stimulation in human subjects. Proc. Natl. Acad. Sci. 83:1140-1144; 1986. Belliveau, J.W. Functional imaging of the brain. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 203. Kwong, K.K.; Belliveau, J.W.; Chesler, D.A.; Goldberg, I.E.; Stern, C.E.; Baker, J.R.; Weisskoff, R.M.; Benson, R.; Poncelet, B.P.; Hoppel, B.E.; Kennedy, D.N.; Turner, R.; Cohen, M.S.; Brady, T. J.; Rosen, B.R. Real time imaging of perfusion change and blood oxygenation change with EPI. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 301. Bandettini, P.A.; Wong, E.C.; Hinks, R.S.; Tikofsky, R.S.; Hyde, J.S. Time-course gradient-echo EPI of localized signal enhancement in the human brain during task activation. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 302. Turner, R.; Jezzard, P.; Wen, H.; Kwong, K.; Le Bihan, D.; Balaban, R. Functional mapping of the human visual cortex at 4 Tesla using deoxygenation contrast EPI. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CASMRM; 1992: 304. Bandettini, P.A.; Wong, E.C.; Hinks, R.S.; Estkowski, L.; Hyde, J.S. Quantification of changes in relaxation rates R2* and R2 in activated brain tissue. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 719. Baker, J.R.; Cohen, M.S.; Stern, C.E.; Kwong, K.K.; Belliveau, J.W.; Rosen, B.R. The effect of slice thickness and echo time on the detection of signal change during echo-planar functional neuroimaging. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1822. Blamire, A.M.; Ogawa, S.; Ugurbil, K.; McCarthy, G.; Ellermann, J.; Hyder, F.; Rattner, Z.; Shulman, R.G. Echo-planar imaging of the activated human visual cortex shows a time delay between stimulus and activation. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1823. DeYoe, E.A.; Neitz, J.; Bandettini, P.A.; Wong, E.C.; Hyde, J.S. Time course of event-related MR signal enhancement in visual and motor cortex. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1824. Rao, SM.; Bandettini, P.A.; Wong, E.C.; Yetkin, F.Z.; Hammeke, T.A. Mueller, W.M.; Goldman, R.S.; Morris, G.L.; Antuono, P.G.; Estkowski, L.D.; Haughton, V.M.; Hyde, J.S. Gradient-echo EPI demonstrates bilateral superior temporal gyrus activation during passive
Functional 2D and 3D MRI at high spatial resolution l L.R. SCHADET
word presentation. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1827. 14. Blamire, A.M.; McCarthy, G.; Gruetter, R.; Rothman, D.L.; Rattner, Z.; Hyder, F.; Shulman, R.G. Echo-planar imaging of the left inferior frontal lobe during word generation. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1834. 15. Frahm, J.; Bruhn, H. ; Merboldt, K.D. ; Hlnicke, W, Dynamic MR imaging of human brain oxygenation during rest and photic stimulation. JMRI 2501-505; 1992. 16. Gore, J.C.; McCarthy, G.; Constable, R.T.; Anderson, A.W.; Kennan, R.P.; Rattner, Z.; Zhong, J. Imaging regional brain activation at 1.5 T using conventional im-
AL.
15
aging techniques. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1826. 17. Frahm, J.; Merboldt, K.D.; Hlnicke, W. Functional MRI of human brain activation at high spatial resolution. Mugn. Reson. Med. 29(1):139-144; 1993. 18. Schad, L.R.; Trost, U.; Knopp, M.V.; Mtiller, E.; Lorenz, W. J. Motor cortex stimulation measured by magnetic resonance imaging on a standard 1.5 T clinical scanner. Magn. Reson. Imaging 11(4):461-464; 1993. 19. Haase, A.; Frahm, J.; Matthaei, D.; Merboldt, K.D.; Haenike, W. Rapid images and NMR movies. In: Book of abstracts: Fourth Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1985: 980-981.
Copyright 0
0730-725X/94 $6.00 + .W 1993 Pergamon Press Ltd.
0 Original Contribution
FUNCTIONAL 2D AND 3D MAGNETIC RESONANCE IMAGING OF MOTOR CORTEX STIMULATION AT HIGH SPATIAL RESOLUTION USING STANDARD 1.5 T IMAGER LOTHAR R. SCHAD,* FREDERIK WENZ,* MICHAEL V. KNOPP,* KLAUS BAUDENDISTEL,* EDGAR M~~LLER,~ AND WALTER J. LORENZ* *Department of Radiology, German Cancer Research Center, Heidelberg, Germany and tSiemens Medical Division, Erlangen, Germany This paper reports the effects of motor cortex stimulation of normal volunteers using conventional MR imaging techniques on standard 1.5 T clinical scanner. Improvement in signal-to-noise (S/N) ratio has been achieved by using a commercially available eye/ear surface coil with a loop of 8.5 cm in diameter. Magnet shimming with all first order coils was performed to the volunteer’s head resulting in a magnetic field homogeneity of about 0.1-0.2 ppm. The imaging technique used was an optimized conventional 2D and 3D, first order flow rephased, gradient-echo sequence (FLASH) with fat-suppression and reduced bandwidth (16-28 Hz/pixel) and TR = 80120 ms, TE = 60 ms, flip angle = 40°, matrix = 128 x 128, FOV = 150-250 mm, slice-thickness = 2-5 mm, NEX = 1, and a total single scan time for one image of about 12-16 s. In the 3D FLASH measurements, a slab of 32 mm thickness with 16 partitions was evaluated. The motor cortex stimulation was achieved by touching each finger to thumb in a sequential, self-paced, and repetitive manner. During stimulation, an increase in signal of order lo-20% was detected in the motor and sensory cortex due to reduced partial volume effects and optimized S/N for the measurements at small voxel size. 3D FLASH imaging at high spatial resolution shows good anatomical correlation of signal increase with gray matter of the motor and sensory cortex. The reported data demonstrate the technical feasibility of functional 2D and 3D MR imaging at high spatial resolution using optimized conventional sequences and equipment.
Keywords: Magnetic resonance imaging, techniques; Functional imaging; Motor cortex stimulation.
INTRODUCTION
effect), while oxygenized haemoglobin is not paramagnetic. On the other hand, regional cerebral blood volume (rCBV) and regional cerebral blood flow (rCBF) increase during stimulation which have been demonstrated by PET measurements showing an increase in rCBF during stimulation of about 5-20% .4 The net effect of regional MR signal change (TT effect modulated by chances in rCBV and rCBF) can be measured by sequences that specifically enhance this Tz effect by using long echo times. At the Eleventh Annual Scientific Meeting of the Society of Magnetic Resonance in Medicine (1992) in Berlin, recently reported studies were presented on completely noninvasive MR imaging of human brain activity based on either blood flow or blood oxygenation changes.5-‘4 Most of the investigators were using echo planar imaging (EPI) methods requiring special hardware equipment. Stimulation
The visualization of brain function and its reaction to external stimuli has always been of great interest both to clinical as well as scientific research. Until recently, this area was the domain of positron emission tomography (PET). New measuring techniques have demonstrated that functional imaging of brain perfusion is also possible with magnetic resonance (MR) imaging.lm3The goal in these studies was to employ the inherent sensitivity of MR to local changes in magnetic susceptibility which are induced by paramagnetic particles like Gadolinium-DTPA, the well known contrast agent in MR. Another even more noninvasive approach is to use the changes and paramagnetic effects of haemoglobin. Deoxygenized haemoglobin is paramagnetic and slightly distorts the magnetic field around it (r,* RECEIVED 2/10/93; ACCEPTED 7/6/93. Address correspondence to PD Dr. L.R.
Schad,
schungsschwerpunkt zentrum, Postfach
For9
Radiologie, Deutsches Krebsforschungs101949, D-69009 Heidelberg, Germany.
10
Magnetic Resonance Imaging 0 Volume 12, Number
measurements on standard scanners using conventional imaging techniques were reported rarely and are essentially preliminary.‘5-‘8 The present study was undertaken to evaluate the feasibility of functional 2D and 3D MR imaging of the human brain at high spatial resolution and with improved signal-to-noise (S/N) ratio on the basis of optimized conventional FLASH (Fast Low Angle Shot) sequences” using standard 1.5 T whole body equipment. MATERIALS
AND METHODS
Instrumentation MR imaging was performed on a 64 MHz Magnetom (Siemens, Erlangen, Germany) superconducting whole body imager. Improvement in S/N of the stimulation measurements has been achieved by using a commercially available eye/ear surface coil with a loop of 8.5 cm in diameter. High loaded Q-values results in an improved S/N ratio and thus high-quality images. Magnet shimming with all first order coils was performed to the volunteer’s head resulting in a magnetic field homogeneity of about 0.1-0.2 ppm. The imaging technique used was an optimized conventional 2D and 3D, first order flow rephased, gradient-echo sequence (FLASH) with fat-suppression and reduced bandwidth (16-28 Hz/pixel) for getting optimal S/N ratio. The sequence parameters used are: TR = 80-120 ms, TE = 60 ms, flip angle = 40”, matrix = 128 x 128, FOV = 150-250 mm, slice-thickness = 2-5 mm, NEX = 1, and a total single scan time for one image of about 12- 16 s. In the 3D FLASH measurements, a slab of 32 mm thickness with 16 partitions was evaluated resulting in a total measuring time for the 3D experiment of about 4 min. All FLASH experiments were anatomically correlated with conventional T1-weighted spin-echo (SE) images. The MR computer (MICRO-VAX, DEC, Maynard, Massachusetts, U.S.A) was connected directly (DEC-NET, DEC, link) to a central computer (VAXWORKSTATION/3600, DEC) where the software program developed by our group for 2D and 3D functional mapping was implemented. After data acquisition, typically 60-100 images were transferred to the central computer. Clinical Study At the time of publication, 20 volunteers were investigated. For orientation, sagittal FLASH images were measured firstly. Based on these sagittal images, 7’,-weighted SE imaging (TR/TE 600/15) has been performed in an oblique plane in between axial and coronal planes with a tilting angle of about 10”. The slices
1, 1994
with the best visualization of the primary motor cortex were chosen for dynamic 2D imaging and high resolution 3D mapping of the stimulation experiments. The motor cortex stimulation was achieved by touching each finger to thumb in a sequential, self-paced, and repetitive manner. RESULTS Figure 1A shows an oblique (axial to coronal: loo) SE image of a normal volunteer with high spatial resolution (FOV = 150 mm, slice-thickness = 2 mm, matrix = 256 x 256) using the surface coil. Figure 1B demonstrates the corresponding result of a dynamic 2D FLASH stimulation experiment (FOV = 150 mm, slicethickness = 2 mm, matrix = 128 x 128) showing a difference image between averages of all stimulation (30 images) and nonstimulation (30 images) images. The stimulation experiment starts with the measurement of 10 control images (without finger movement) followed by 10 corresponding stimulation images (with finger movement of the right hand). This stimulus alternation was repeated for three cycles. Comparing the difference image with the T, -weighted SE image a signal increase produced by this stimulation (touching each finger to thumb) occurred in areas predominantly occupied by the grey matter of primary left motor and sensory cortex. Figure 2A reflects two representative signal-time curves in the primary motor cortex (region of interest mean values including about 20 pixels) and the sensory cortex during the stimulation experiments. The data show a significant increase in signal of order lo-20%. The effects in signal change during stimulation can be seen in one pixel only with an increase in signal of order 20% (Fig. 2B). Figure 2C represents the results of the same stimulation experiment (signal-time curve in the primary motor cortex), but of another volunteer where a clear linear signal increase during stimulation was observed. Figure 3 shows eight central slices out of a total of 16 slices as a typical example of a 3D FLASH stimulation experiment of a normal volunteer using the surface coil. Thus, oblique (axial to coronal: 10’) SE images (FOV = 200 mm, slice-thickness = 2 mm, matrix = 256 x 256) were overlayed by the corresponding difference images (green color) of the 3D FLASH stimulation experiment (FOV = 200 mm, slab-thickness = 32 mm/16 partitions, matrix = 128 x 128). In particular, the 3D stimulation experiment starts with a 3D FLASH measurement (16 slices in 4 min) without finger movement, followed by a corresponding 3D FLASH measurement (16 slices in 4 min) with finger movement of the right hand. This stimulus alternation was repeated for eight cycles. The difference images (Fig. 3,
Functional 2D and 3D MRI at high spatial resolution 0 L.R.
(4
SCHAD
ET AL.
11
@I
Fig. 1. 2D stimulation experiment. (A) Oblique (axial to coronal: 10”) SE image (TRITE 600/15) of a normal volunteer with high spatial resolution (FOV = 150 mm, TH = 2 mm, matrix = 256 x 256) using the surface coil. (B) Difference image (FOV = 150 mm, TH = 2 mm, matrix = 128 x 128) between averages of all stimulation (30 images) and nonstimulation (30 images) images showing signal increase produced by the stimulation (touching each finger to thumb) occurred in areas predominantly occupied by the grey matter of primary left motor and sensory cortex.
green overlay color) were calculated by averaging of all 3D stimulation images (8 images at every slice position) and nonstimulation images, and subtraction of each other. The 3D FLASH stimulation measurement at high spatial resolution demonstrates again that signal increase produced by this kind of stimulation (touching each finger to thumb) occurred in areas predominantly occupied by the grey matter of primary left motor and sensory cortex. DISCUSSION As expected, stimulation experiments at high spatial resolution show a clear enhancement of the observed stimulation effect (about 20%) and an improvement in detectability. As seen from the head coil experiments at larger voxel size,” an increase in signal intensity due to larger voxels does not lead directly to improvements in the detectability of signal change because of partial volume averaging with nonstimulated areas of the brain. Of course, reduced voxel size lead to a reduction of S/N. Improvement in S/N can be achieved by using surface coils with high loaded Q-values and by optimization of the measuring sequences. Reduction in the bandwidth to about 20 Hz/pixel does significantly increase the
S/N. On the other hand, chemical shift artefacts caused by the fat mainly from the outer side of the head lead to a shift between fat and water images of about 10 pixels in read-out direction. This chemical shift artefact can shift parts of the fat image onto areas of motor or sensory cortex and thus obscure the stimulation effects. To overcome this problem we used standard fat suppression in the FLASH sequences. In our experience, an effective fat suppression can be reached for the head coil and surface coil experiments at a magnetic field homogeneity of about 0.1-0.2 ppm. 2D and 3D FLASH stimulation experiments at high spatial resolution compared with Tl-weighted SE images show a clear anatomical correlation between signal increase and grey matter of primary motor cortex and sensory cortex. Although the 3D FLASH experiments (Fig. 3) show activation along each sulcus, the major effects are confined to the outer layers of grey matter along motor and sensory cortex. The additional highlighted regions may arise from pulsation of cerebrospinal fluid or may even be caused by flow effects and vascular pulsation artefacts of large vessels, that is, superior sagittal sinus. However, yet limited in the number of volunteers, the available data reveal a remarkable interindividual reproducibility of the expected
Magnetic Resonance Imaging 0 Volume 12, Number 1, 1994
12
600
800 time
[set]
surface coil 1 DiXd
25C
200
600
400
200
0
800 time
[set]
W 130’-
7
.
surface coil 20pki 120.
ti B # #.
ilO0
rl, 0
#
l.,
.o -0
l 0
stirmalinl
800
600
time
Fig. 2. 2D stimulation
[eec]
dynamic. (A) Typical signal-time curves of the primary motor cortex and sensory cortex (region of interest mean values including about 20 pixels) showing a significant increase in’signal of order lo-20% during stimulation. (B) Effects in signal change can be seen in one pixel only with an increase in signal of about 20% during stimulation. (C) Results of the same stimulation experiment (signal-time curve in the primary motor cortex) but of another volunteer where a clear linear signal increase during stimulation was observed.
0)
09
:E)
Fig. 3. 3D stimulation experiment. (A-H) Eight central slices out of a total of 16 slices of a typical 3D FLASH stimulation experiment of a normal volunteer using the surface coil (FOV = 200 mm, SLAB = 32 mm/16 PART, matrix = 128 x 128). Oblique (axial to coronal: 10”) SE images were overlayed by the corresponding difference images (green color) of the 3D FLASH stimulation experiment showing again a clear anatomical correlation between signal increase produced by this kind of stimulation (touching each finger to thumb) and areas predominantly occupied by the grey matter of primary left motor and sensory cortex. A maximum intensity projection (MIP) of the stimulated areas seems to be possible.
(4
W
14
Magnetic Resonance Imaging 0 Volume 12, Number 1, 1994
activation maps. The observed increase in signal intensity in the activated areas of the brain is potentially explained qualitatively by findings previously confirmed by PET observations,4 that during local brain stimulation, oxygen delivery to the activated region exceeds metabolic need. The overabundance of oxygen-rich blood leads to a decrease in the local oxygen extraction fraction, a local increase in O2 utilization, and a local decrease in the concentration of deoxyhemoglobin. Thereby deoxyhemoglobin acts as an effective endogenous contrast agent where signal variation during stimulation is caused by differences in the magnetic susceptibility of blood induced by the presence of paramagnetic deoxyhemoglobin in red cells. A decrease in deoxyhemoglobin concentration decreases the vesseltissue susceptibility differential, allowing increased spin coherence and thus increased signal in gradient-echo imaging (increase of T,*). Our data (20 volunteers) show different signal behaviour during stimulation: most of the volunteers showed a more or less constant enhanced signal during stimulation, but in some cases a clear linear signal increase was observed (Fig. 2C). At the moment, the neurophysiological explanation of this different signal behaviour is unclear and needs some further clinical investigations completed by PET experiments. In conclusion, our data demonstrate the technical feasibility of 2D and 3D functional MR imaging at high spatial resolution on standard clinical 1.5 T scanner using optimized gradient-echo sequences and coil equipment. 2D time course gradient-echo imaging is a powerful new noninvasive tool for assessment of regional cerebral activation with high temporal and spatial resolution. 3D functional imaging is another powerful tool to image the stimulation effects of the whole brain cortex and a maximum intensity projection (MIP), the well known projection method in MR angiography, seems to be possible.
5.
6.
7.
8.
9.
10.
11.
REFERENCES Ogawa, S.; Lee, T.M. ; Nayak, A.S.; Glynn, P. Oxygenation-sensitive contrast in magnetic resonance image of rodent brain at high magnetic fields. Magn. Reson. Med. 14:68-78; 1990. Rosen, B.R.; Belliveau, J.W.; Vevea, J.M.; Brady, T.J. Perfusion imaging with NMR contrast agents. Magn. Reson. Med. 14:249-265; 1990. Belliveau, J.W.; Kennedy, D.N.; McKinstry, R.C.; Buchbinder, B.R.; Weisskoff, R.M.; Cohen, M.S.; Vevea, J.M.; Brady, T.J.; Rosen, B.R. Functional mapping of the human visual cortex by magnetic resonance imaging. Science 254:716-719; 1991. Fox, P.T.; Raichle, M.E. Focal physiological uncoupling
12.
13.
of cerebral blood flow and oxidative metabolism during somatosensory stimulation in human subjects. Proc. Natl. Acad. Sci. 83:1140-1144; 1986. Belliveau, J.W. Functional imaging of the brain. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 203. Kwong, K.K.; Belliveau, J.W.; Chesler, D.A.; Goldberg, I.E.; Stern, C.E.; Baker, J.R.; Weisskoff, R.M.; Benson, R.; Poncelet, B.P.; Hoppel, B.E.; Kennedy, D.N.; Turner, R.; Cohen, M.S.; Brady, T. J.; Rosen, B.R. Real time imaging of perfusion change and blood oxygenation change with EPI. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 301. Bandettini, P.A.; Wong, E.C.; Hinks, R.S.; Tikofsky, R.S.; Hyde, J.S. Time-course gradient-echo EPI of localized signal enhancement in the human brain during task activation. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 302. Turner, R.; Jezzard, P.; Wen, H.; Kwong, K.; Le Bihan, D.; Balaban, R. Functional mapping of the human visual cortex at 4 Tesla using deoxygenation contrast EPI. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CASMRM; 1992: 304. Bandettini, P.A.; Wong, E.C.; Hinks, R.S.; Estkowski, L.; Hyde, J.S. Quantification of changes in relaxation rates R2* and R2 in activated brain tissue. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 719. Baker, J.R.; Cohen, M.S.; Stern, C.E.; Kwong, K.K.; Belliveau, J.W.; Rosen, B.R. The effect of slice thickness and echo time on the detection of signal change during echo-planar functional neuroimaging. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1822. Blamire, A.M.; Ogawa, S.; Ugurbil, K.; McCarthy, G.; Ellermann, J.; Hyder, F.; Rattner, Z.; Shulman, R.G. Echo-planar imaging of the activated human visual cortex shows a time delay between stimulus and activation. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1823. DeYoe, E.A.; Neitz, J.; Bandettini, P.A.; Wong, E.C.; Hyde, J.S. Time course of event-related MR signal enhancement in visual and motor cortex. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1824. Rao, SM.; Bandettini, P.A.; Wong, E.C.; Yetkin, F.Z.; Hammeke, T.A. Mueller, W.M.; Goldman, R.S.; Morris, G.L.; Antuono, P.G.; Estkowski, L.D.; Haughton, V.M.; Hyde, J.S. Gradient-echo EPI demonstrates bilateral superior temporal gyrus activation during passive
Functional 2D and 3D MRI at high spatial resolution l L.R. SCHADET
word presentation. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1827. 14. Blamire, A.M.; McCarthy, G.; Gruetter, R.; Rothman, D.L.; Rattner, Z.; Hyder, F.; Shulman, R.G. Echo-planar imaging of the left inferior frontal lobe during word generation. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1834. 15. Frahm, J.; Bruhn, H. ; Merboldt, K.D. ; Hlnicke, W, Dynamic MR imaging of human brain oxygenation during rest and photic stimulation. JMRI 2501-505; 1992. 16. Gore, J.C.; McCarthy, G.; Constable, R.T.; Anderson, A.W.; Kennan, R.P.; Rattner, Z.; Zhong, J. Imaging regional brain activation at 1.5 T using conventional im-
AL.
15
aging techniques. In: Book of abstracts: Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1992: 1826. 17. Frahm, J.; Merboldt, K.D.; Hlnicke, W. Functional MRI of human brain activation at high spatial resolution. Mugn. Reson. Med. 29(1):139-144; 1993. 18. Schad, L.R.; Trost, U.; Knopp, M.V.; Mtiller, E.; Lorenz, W. J. Motor cortex stimulation measured by magnetic resonance imaging on a standard 1.5 T clinical scanner. Magn. Reson. Imaging 11(4):461-464; 1993. 19. Haase, A.; Frahm, J.; Matthaei, D.; Merboldt, K.D.; Haenike, W. Rapid images and NMR movies. In: Book of abstracts: Fourth Annual Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, CA:SMRM; 1985: 980-981.