- Email: [email protected]
Dual representation of the hand in the cerebellum: activation with voluntary and passive finger movement
Available online at www.sciencedirect.com R
NeuroImage 18 (2003) 670 – 674
www.elsevier.com/locate/ynimg
Dual representation of the hand in the cerebellum: activation with voluntary and passive finger movement Gary W. Thickbroom,a,* Michelle L. Byrnes,a and Frank L. Mastagliaa,b a
Centre for Neuromuscular and Neurological Disorders, University of Western Australia, QEII Medical Centre, Nedlands, WA 6009, Australia b Department of Medicine, University of Western Australia, QEII Medical Centre, Nedlands, WA 6009, Australia Received 10 June 2002; revised 26 September 2002; accepted 14 October 2002
Abstract Early electrophysiological studies during sensory stimulation in the anesthetized cat and more recent functional imaging studies during voluntary movement in humans have provided evidence for two separate representations of the body in the anterior and posterior lobes of the cerebellum; however, the functional role of these body maps in motor and sensory processing is not known. The aims of the present study were to determine whether this dual representation is also present during passive movement, and to compare the pattern of activation with that obtained during voluntary movement. Functional MRI measurements were undertaken in 14 subjects who performed right index finger flexion and extension movements at ⬃1 Hz, or had their finger moved passively at the same rate and through the same angle using a pneumatic device. During passive movement, dual activation was detected in the ipsilateral cerebellum, in the anterior lobe, and in the posterior lobe. A similar pattern of activation was observed during voluntary movement; however, the overall magnitude was about doubled. These data provide evidence for a dual ipsilateral representation of the hand in the rostral and caudal cerebellar cortex during passive as well as voluntary movements, with the rostral representation being the dominant one, and indicate that both of these areas are involved in kinesthetic sensory and motor processing. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Cerebellum; Hand representation; Active; Passive; fMRI
Introduction Early electrophysiological studies in the anesthetized cat and monkey provided evidence that sensory stimulation can activate cells in the cerebellar cortex in the absence of voluntary movement, and showed that there were two separate representations of the body in the anterior and the posterior lobes (Combs, 1954; Snider and Eldred, 1952). However, this dual body map was lost in the awake cat, in which a more diffuse projection pattern was found. More recent functional
* Corresponding author. Brain Research Laboratory, Centre for Neuromuscular and Neurological Disorders, University of Western Australia, Queen Elizabeth II Medical Centre, Nedlands WA 6009, Australia. Fax: ⫹61-8-9346-3487. E-mail address: [email protected] (G.W. Thickbroom).
imaging studies in the awake human have also identified a dual hand representation in the anterior and posterior lobes of the cerebellum during voluntary movement (Grodd et al., 2001; Rijntjes et al., 1999). These studies have not investigated the patterns of activation during passive limb movement, and it is not known whether there is a similar dual representation for afferent inputs to the cerebellum, as in the cat and monkey. If this is the case, the question that arises is whether the two areas have comparable roles in motor and sensory processing, or whether there are differences in the patterns of activity in these areas during voluntary movement and passive kinesthetic sensory stimulation. The aims of the present study were, therefore, to determine whether a dual representation can also be demonstrated with passive movement, and to compare the patterns and degree of cerebellar activation with kinematically comparable active and passive limb movement.
1053-8119/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S1053-8119(02)00055-1
G.W. Thickbroom et al. / NeuroImage 18 (2003) 670 – 674
671
Materials and methods
Analysis
Subjects
Functional images were generated by subtraction analysis using Student’s t test comparing the rest and active conditions on a voxel-by-voxel basis. Data analysis was carried out without employing spatial smoothing, omitting the first two image sets in each block to allow for the time course of the hemodynamic response (leaving 20 data sets in the active and rest conditions), and with a significance level of P ⬍ 0.005 (one tail; t ⬎ 2.7; 38 df; Thickbroom et al., 2000). Analysis was performed immediately following each acquisition run, and runs that were degraded due to motion artifact were repeated. The total number of activated voxels was calculated for regions of interest drawn to encompass areas of activation in the ipsilateral cerebellum, contralateral sensorimotor cortex, and supplementary motor area.
The study had the approval of the Ethics Committee of the University of Western Australia. Fourteen healthy righthanded subjects (22–53 years of age, 6F) gave informed consent to participate in the study. Image acquisition Functional imaging was carried out on a 1.5 T Siemens Magnetom Vision Plus scanner equipped with gradient overdrive and echo-planar imaging (EPI) capabilities. Imaging was performed using a standard head coil with 256⫻256-mm field of view and a 64⫻64 image matrix (4⫻4-mm in-plane voxel size). Subjects lay in the supine position with the head held steady using a bitemporal clamp. The anterior and posterior commissures (AC, PC) were identified from a set of 3 sagittal slices. Thirty-four 3-mmthick axial slices with a 1-mm spacing which lay parallel to the AC–PC line, the first slice tangential to the superior surface of the brain, were selected for functional imaging, resulting in sampling of the whole brain. Three-dimensional iso-voxel anatomical reference images (1 mm3, 256 mm FOV) were acquired using an MPRAGE sequence. Functional imaging was carried out using a blood oxygen-leveldependent (BOLD) gradient-recalled echo-planar sequence (90 degree pulse, TE ⫽ 66 ms). Six sets of images were collected at 4-s intervals during alternating 24-s periods of rest and activation. Each rest–activation cycle was repeated five times, resulting in a total of 60 images per slice for each experimental run. Experimental procedure Two tasks were employed in separate runs. In the first, the index finger of the right hand was flexed and extended passively at ⬃1 Hz using a purpose-built pneumatically driven device with a range of motion of 45 degrees. For the voluntary task, the finger remained in the passive device, and subjects were asked to move their index finger at the same rate and through the same range as for passive movement. The mechanical limits of the device enabled subjects to perform the same range of movement. The 1-Hz rate was readily estimated by subjects, and the actual rate was recorded by an observer who counted the number of movements in each run. Subjects were closely observed during the scanning procedure, and were instructed to remain completely relaxed during the passive movement protocol and not to assist with the movement or imagine the movement of the finger. In addition, EMG recordings were made from the flexor digitorum sublimis and extensor digitorum communis muscles outside the scanner to confirm that subjects were not providing voluntary assistance during the passive movement.
Results Cerebellar activation During voluntary movement, activation was observed in two separate regions of the ipsilateral cerebellar hemisphere, one located rostrally in the anterior lobe, and the other caudally in the posterior lobe (Figs. 1A–1C). These regions correspond to Schmahmann hemispheric lobules V/VI (anterior region) and XIIIB/IX (posterior region) (Schmahmann et al., 1999), and are consistent with previous reports of the pattern of cerebellar activation during voluntary movement (Grodd et al., 2001; Rijntjes et al., 1999). These two pairs of lobules are separated by the primary and secondary cerebellar fissures, respectively (Schmahmann et al., 1999). The degree of activation (number of activated voxels) was greater in the anterior region compared with the posterior region (Table 1; Fig. 2). Activation was also observed in the ipsilateral cerebellum during passive finger movement in 12 of the 14 subjects. This activation was observed in the same two regions of the cerebellar hemisphere as with voluntary finger movement (Figs. 1D and 1E), and was greater for the anterior region than the posterior region (Table 1; Fig. 2). The overall degree of activation was proportionally greater during voluntary movement than passive movement, being approximately twofold for the anterior region compared to the posterior region. The relative degree of activation between the anterior and the posterior regions during voluntary movement (ratio 24.6:9.6 ⫽ 2.6) was similar to that during passive movement (9.6:4.2 ⫽ 2.3). There were approximately 32 movements performed in each 24-s activation task block for both voluntary and passive movement. Activation was not observed in the contralateral cerebellum during voluntary or passive movement.
672
G.W. Thickbroom et al. / NeuroImage 18 (2003) 670 – 674
Fig. 1. Dual regions of cerebellar activation during voluntary right index finger movement shown for one subject (A) in the transverse plane (two noncontiguous slices, see also (D)), (B) sagittal plane, and (C) coronal plane. Similar dual regions of activation were observed during voluntary (D) and passive (E) finger movement (10 contiguous transverse slices, second subject).
Cerebral cortical activation In all subjects, voluntary movement resulted in activation of the contralateral primary sensorimotor cortex (SM1) and supplementary motor area (SMA), in keeping with the expected pattern of cortical activation during a finger-tapping task. Likewise with passive finger movement, a similar pattern of cortical activation was obtained, but the degree of activation was less than that during the voluntary movement task in both the SM1 and the SMA (Table 1; Fig. 2). Proportionally, the relative pattern of activation was the same for both regions during voluntary and passive movement, with almost double the activation during voluntary movement for both SM1 (ratio 40.8:26.4 ⫽ 1.5) and SMA (10.8:6.2 ⫽ 1.7).
Discussion
stimulation, or passive movement in some studies (Casey et al., 1996; Mima et al., 1999; Seitz and Roland, 1992; Tempel and Perlmutter, 1992), whereas in other studies activation at times equal to that recorded during voluntary movements has been described with passive movement or sensory stimulation of the upper limb (Jueptner et al., 1997; Jueptner and Weiller, 1998; Nitschke et al., 1998). We found cerebellar activation during passive finger movement; however, the overall level of activation was lower than with a comparable voluntary movement. As the sensory input during passive limb movement presumably involves a combination of tactile, joint, and muscle afferent activity, it is Table 1 Number of activated voxels (group mean ⫾ SD) in the cerebellum (anterior and posterior regions) and cerebral cortex (primary sensorimotor cortex and supplementary motor area) for voluntary and passive movement Cerebellum
There has been conflicting evidence for the presence of cerebellar activation with sensory stimulation in previous human functional imaging studies. Cerebellar activation was not found during vibro-tactile stimulation, noxious
Voluntary Passive
Cerebrum
Anterior
Posterior
SM1
SMA
24.6 ⫾ 2.8 9.6 ⫾ 2.3
9.6 ⫾ 2.4 4.2 ⫾ 1.3
40.8 ⫾ 6.4 26.4 ⫾ 4
10.8 ⫾ 2.2 6.2 ⫾ 1.6
G.W. Thickbroom et al. / NeuroImage 18 (2003) 670 – 674
673
Fig. 2. Group mean data comparing the number of activated voxels in the contralateral sensorimotor cortex (SM1), midline supplementary motor area (SMA) and anterior and posterior ipsilateral cerebellar hemispheres during voluntary and passive movement (mean ⫾ standard error).
possible that the cerebellum is involved in discriminating and integrating these multiple channels of sensory information. Studies that have sought to preferentially activate a single sensory modality, such as vibration alone, have generally been unsuccessful in eliciting cerebellar activation (Casey et al., 1996; Mima et al., 1999 Seitz and Roland, 1992; Tempel and Perlmutter, 1992). Thus, integration of sensory information in the cerebellum from a range of receptors may be the basis for the present observations, and while there was no requirement for overt sensory discrimination, a form of sensory pattern processing may have occurred, resulting in a pattern of activation similar to that during a comparable voluntary movement. Although it was not technically possible to monitor EMG activity in the forearm muscles during the fMRI studies, a number of measures were taken to ensure that there was no voluntary contribution by the subjects during the passive movement studies. These included the use of only experienced subjects, and a preliminary instruction and training session. In addition, in a number of subjects EMG recordings from the forearm muscles outside the scanner failed to show motor unit recruitment during the passive finger movement. While motor imagery during the passive protocol is a further consideration, previous studies of motor imagery have indicated that this results in only weak cerebellar activation and is unlikely to account for the degree of activation observed in the present study (Jueptner et al., 1997; Lotze et al., 1999). The possibility was considered that the dual pattern of activation found might be a consequence of the organization of efferent and afferent cerebellar pathways. For example, could the posterior hand representation have a preferential role in processing kinesthetic spinal inputs via the inferior
cerebellar peduncle? Against this is the fact that these fibers are known to terminate widely throughout the cerebellum, overlapping in both the anterior and posterior lobes, as do the cerebral cortical afferents traveling via the middle peduncle (Rothwell, 1994). Consequently, it appears that sensory afferents are widely distributed in the cerebellum, in keeping with the generally uniform neural organization of this organ, and that the dual hand representation cannot be readily explained on purely anatomical grounds. The observation that the dual cerebellar representation in the anesthetized cat is not present in the awake animal (Combs, 1954) is in keeping with activation of diffuse afferent projections from specific body parts throughout the cerebellum in the awake animal. It has been suggested that during anesthesia only the most direct afferent and efferent pathways remain functional, and that this leads to a simplified somatotopy (Rothwell, 1994). A possible explanation for our finding of a dual representation in the awake human with functional imaging is that functional MRI may detect activation of only the most direct afferent and efferent pathways involving the fewest synapses and leading to simultaneous activation of large populations of cells in a way that yields a significant BOLD response. Thus some caution is warranted in ascribing functional significance to the dual representation in the human, and it remains possible that cerebellar processing is in fact undertaken using more widespread and less spatially defined networks. This is also supported by the difficulty in demonstrating a clear somatotopy within the cerebellum comparable to that readily identified in primary motor and sensory cortex (Grodd et al., 2001; Nitschke et al., 1996; Rijntjes et al., 1999). It is possible that the cytoarchitectonic uniformity of the cerebellum and the diffuse patterns of afferent fiber terminations
674
G.W. Thickbroom et al. / NeuroImage 18 (2003) 670 – 674
in the cerebellar cortex mean that the cerebellum is not preferentially structured to process information in a segregated somatotopic manner. Previous studies of voluntary and passive movement have shown that the pattern of activation in cerebral cortex is similar during both of these tasks (Mima et al., 1999; Weiller et al., 1996; Yetkin et al., 1995; Mima et al., 1999). We observed a relative pattern of activation in the sensorimotor cortex and SMA similar to that in the cerebellum during voluntary and passive movement, the degree of activation in all regions being approximately halved during passive movement. This suggests that a similar global network is activated during both of these tasks, and further highlights the close association between kinesthetic sensory and motor processing throughout the brain at both cortical and subcortical levels. The greater degree of overall activation during voluntary compared with passive movement could be explained by increased neural activation or recruitment as a result of sensory and motor integration during voluntary movement. However, although we attempted to do everything possible to make the voluntary and passive movements the same in all respects, there are likely to be differences in the patterns of afferent inputs in the two situations, and we cannot exclude the possibility that the differences we have observed reflect different levels of afferent inputs during voluntary and passive movement. As well, differences in attention during voluntary movement compared to passive movement could affect level of activation in these two conditions. In conclusion, we have shown that it is possible to demonstrate cerebellar activation with passive movement of a single digit using fMRI and that there is a dual pattern of activation as in the case of voluntary movement. While the functional significance of the dual hand representation in the anterior and posterior lobes of the cerebellum remains uncertain, the present findings indicate that both areas are involved in sensory and motor processing.
Acknowledgments The authors are grateful to Dr. S. Ghosh for comments and helpful discussion. Dr. M. Fallon, Dr. S. Davis, and Mr. I. Morris, Department of Radiology, MRI Unit, Sir Charles Gairdner Hospital, and radiographers from the MRI Unit are thanked for support and assistance with the imaging studies. Prof. L.A. Cala and members of the Australian Research Centre for Medical Engineering are thanked for assistance with the passive movement device.
References Casey, K.L., Minoshima, S., Morrow, T.J., Koeppe, R.A., 1996. Comparison of human cerebral activation pattern during cutaneous warmth, heat pain, and deep cold pain. J. Neurophysiol. 76, 571–581. Combs, C.M., 1954. Electro-anatomical study of cerebellar localisation. Stimulation of various afferents. J. Neurosci. 17, 123–143. Grodd, W., Hulsmann, E., Lotze, M., Wildgruber, D., Erb, M., 2001. Sensorimotor mapping of the human cerebellum: fMRI evidence of somatotopic organisation. Human Brain Mapp. 13, 55–73. Jueptner, M., Ottinger, S., Fellows, S.J., Adamschewski, J., Flerich, L., Muller, S.P., Diener, H.C., Thilmann, A.F., Weiller, C., 1997. The relevance of sensory input for the cerebellar control of movements. NeuroImage 5, 41– 48. Jueptner, M., Weiller, C., 1998. A review of differences between basal ganglia and cerebellar control of movements as revealed by functional imaging studies. Brain 121, 1437–1449. Lotze, M., Montoya, P., Erb, M., Hulsmann, E., Flor, H., Klose, U., Birbaumer, N., Grodd, W., 1999. Activation of cortical and cerebellar motor areas during executed and imagined hand movements: an fMRI study. J. Cogn. Neurosci. 11, 491–501. Mima, T., Sadato, N., Yazawa, S., Hanakawa, T., Fukuyama, H., Yonekura, Y., Shibasaki, H., 1999. Brain structures related to active and passive finger movements in man. Brain 122, 1989 –1997. Nitschke, M.F., Hahn, C., Melchert, U.H., Handels, H., Wessel, K., 1998. Activation of the cerebellum by sensory finger stimulation and by finger opposition movements—a functional magnetic resonance imaging study. J. Neuroimaging 8, 127–131. Nitschke, M.F., Kleinschmidt, A., Wessel, K., Frahm, J., 1996. Somatotopic motor representation in the human anterior cerebellum. A highresolution functional MRI study. Brain 119, 1023–1029. Rijntjes, M., Buechel, C., Kiebel, S., Weiller, C., 1999. Multiple somatotopic representations in the human cerebellum. NeuroReport 10, 3653– 3658. Rothwell, J.C., 1994. Control of Human Voluntary Movement. Chapman & Hall, London. Schmahmann, J.D., Doyon, J., McDonald, D., Holmes, C., Lavoie, K., Hurwitz, A.S., Kabani, N., Toga, A., Evans, A., Petrides, M., 1999. Three-dimensional MRI atlas of the human cerebellum in proportional sterotaxic space. NeuroImage 10, 233–260. Seitz, R.J., Roland, P.E., 1992. Vibratory stimulation increases and decreases the regional cerebral blood flow and oxidative metabolism: a positron emission tomography (PET) study. Acta Neurol. Scand. 86, 60 – 67. Snider, R.S., Eldred, E., 1952. Cerebrocerebellar relationships in the monkey. J. Neurophysiol. 15, 27– 40. Tempel, L.W., Perlmutter, J.S., 1992. Vibration-induced regional cerebral blood flow responses in normal ageing. J. Cerebral Blood Flow Metab. 12, 554 –561. Thickbroom, G.W., Byrnes, M.L., Sacco, P., Ghosh, S., Morris, I.T., Mastaglia, F.L., 2000. The role of the supplementary motor area in externally timed movement: the influence of predictability of movement timing. Brain Res. 874, 233–241. Weiller, C., Juptner, M., Fellows, S., Rijntjes, M., Leonhardt, G., Kiebel, S., Muller, S., Diener, H.C., Thilmann, A.F., 1996. Brain representation of active and passive movements. NeuroImage 4, 105–110. Yetkin, F.Z., Muellwe, W.M., Hammeke, T.A., Morris, G.L., Haughton, V.M., 1995. Functional magnetic resonance imaging mapping of sensorimotor cortex with tactile stimulation. Neurosurgery 36, 921– 925.
NeuroImage 18 (2003) 670 – 674
www.elsevier.com/locate/ynimg
Dual representation of the hand in the cerebellum: activation with voluntary and passive finger movement Gary W. Thickbroom,a,* Michelle L. Byrnes,a and Frank L. Mastagliaa,b a
Centre for Neuromuscular and Neurological Disorders, University of Western Australia, QEII Medical Centre, Nedlands, WA 6009, Australia b Department of Medicine, University of Western Australia, QEII Medical Centre, Nedlands, WA 6009, Australia Received 10 June 2002; revised 26 September 2002; accepted 14 October 2002
Abstract Early electrophysiological studies during sensory stimulation in the anesthetized cat and more recent functional imaging studies during voluntary movement in humans have provided evidence for two separate representations of the body in the anterior and posterior lobes of the cerebellum; however, the functional role of these body maps in motor and sensory processing is not known. The aims of the present study were to determine whether this dual representation is also present during passive movement, and to compare the pattern of activation with that obtained during voluntary movement. Functional MRI measurements were undertaken in 14 subjects who performed right index finger flexion and extension movements at ⬃1 Hz, or had their finger moved passively at the same rate and through the same angle using a pneumatic device. During passive movement, dual activation was detected in the ipsilateral cerebellum, in the anterior lobe, and in the posterior lobe. A similar pattern of activation was observed during voluntary movement; however, the overall magnitude was about doubled. These data provide evidence for a dual ipsilateral representation of the hand in the rostral and caudal cerebellar cortex during passive as well as voluntary movements, with the rostral representation being the dominant one, and indicate that both of these areas are involved in kinesthetic sensory and motor processing. © 2003 Elsevier Science (USA). All rights reserved. Keywords: Cerebellum; Hand representation; Active; Passive; fMRI
Introduction Early electrophysiological studies in the anesthetized cat and monkey provided evidence that sensory stimulation can activate cells in the cerebellar cortex in the absence of voluntary movement, and showed that there were two separate representations of the body in the anterior and the posterior lobes (Combs, 1954; Snider and Eldred, 1952). However, this dual body map was lost in the awake cat, in which a more diffuse projection pattern was found. More recent functional
* Corresponding author. Brain Research Laboratory, Centre for Neuromuscular and Neurological Disorders, University of Western Australia, Queen Elizabeth II Medical Centre, Nedlands WA 6009, Australia. Fax: ⫹61-8-9346-3487. E-mail address: [email protected] (G.W. Thickbroom).
imaging studies in the awake human have also identified a dual hand representation in the anterior and posterior lobes of the cerebellum during voluntary movement (Grodd et al., 2001; Rijntjes et al., 1999). These studies have not investigated the patterns of activation during passive limb movement, and it is not known whether there is a similar dual representation for afferent inputs to the cerebellum, as in the cat and monkey. If this is the case, the question that arises is whether the two areas have comparable roles in motor and sensory processing, or whether there are differences in the patterns of activity in these areas during voluntary movement and passive kinesthetic sensory stimulation. The aims of the present study were, therefore, to determine whether a dual representation can also be demonstrated with passive movement, and to compare the patterns and degree of cerebellar activation with kinematically comparable active and passive limb movement.
1053-8119/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S1053-8119(02)00055-1
G.W. Thickbroom et al. / NeuroImage 18 (2003) 670 – 674
671
Materials and methods
Analysis
Subjects
Functional images were generated by subtraction analysis using Student’s t test comparing the rest and active conditions on a voxel-by-voxel basis. Data analysis was carried out without employing spatial smoothing, omitting the first two image sets in each block to allow for the time course of the hemodynamic response (leaving 20 data sets in the active and rest conditions), and with a significance level of P ⬍ 0.005 (one tail; t ⬎ 2.7; 38 df; Thickbroom et al., 2000). Analysis was performed immediately following each acquisition run, and runs that were degraded due to motion artifact were repeated. The total number of activated voxels was calculated for regions of interest drawn to encompass areas of activation in the ipsilateral cerebellum, contralateral sensorimotor cortex, and supplementary motor area.
The study had the approval of the Ethics Committee of the University of Western Australia. Fourteen healthy righthanded subjects (22–53 years of age, 6F) gave informed consent to participate in the study. Image acquisition Functional imaging was carried out on a 1.5 T Siemens Magnetom Vision Plus scanner equipped with gradient overdrive and echo-planar imaging (EPI) capabilities. Imaging was performed using a standard head coil with 256⫻256-mm field of view and a 64⫻64 image matrix (4⫻4-mm in-plane voxel size). Subjects lay in the supine position with the head held steady using a bitemporal clamp. The anterior and posterior commissures (AC, PC) were identified from a set of 3 sagittal slices. Thirty-four 3-mmthick axial slices with a 1-mm spacing which lay parallel to the AC–PC line, the first slice tangential to the superior surface of the brain, were selected for functional imaging, resulting in sampling of the whole brain. Three-dimensional iso-voxel anatomical reference images (1 mm3, 256 mm FOV) were acquired using an MPRAGE sequence. Functional imaging was carried out using a blood oxygen-leveldependent (BOLD) gradient-recalled echo-planar sequence (90 degree pulse, TE ⫽ 66 ms). Six sets of images were collected at 4-s intervals during alternating 24-s periods of rest and activation. Each rest–activation cycle was repeated five times, resulting in a total of 60 images per slice for each experimental run. Experimental procedure Two tasks were employed in separate runs. In the first, the index finger of the right hand was flexed and extended passively at ⬃1 Hz using a purpose-built pneumatically driven device with a range of motion of 45 degrees. For the voluntary task, the finger remained in the passive device, and subjects were asked to move their index finger at the same rate and through the same range as for passive movement. The mechanical limits of the device enabled subjects to perform the same range of movement. The 1-Hz rate was readily estimated by subjects, and the actual rate was recorded by an observer who counted the number of movements in each run. Subjects were closely observed during the scanning procedure, and were instructed to remain completely relaxed during the passive movement protocol and not to assist with the movement or imagine the movement of the finger. In addition, EMG recordings were made from the flexor digitorum sublimis and extensor digitorum communis muscles outside the scanner to confirm that subjects were not providing voluntary assistance during the passive movement.
Results Cerebellar activation During voluntary movement, activation was observed in two separate regions of the ipsilateral cerebellar hemisphere, one located rostrally in the anterior lobe, and the other caudally in the posterior lobe (Figs. 1A–1C). These regions correspond to Schmahmann hemispheric lobules V/VI (anterior region) and XIIIB/IX (posterior region) (Schmahmann et al., 1999), and are consistent with previous reports of the pattern of cerebellar activation during voluntary movement (Grodd et al., 2001; Rijntjes et al., 1999). These two pairs of lobules are separated by the primary and secondary cerebellar fissures, respectively (Schmahmann et al., 1999). The degree of activation (number of activated voxels) was greater in the anterior region compared with the posterior region (Table 1; Fig. 2). Activation was also observed in the ipsilateral cerebellum during passive finger movement in 12 of the 14 subjects. This activation was observed in the same two regions of the cerebellar hemisphere as with voluntary finger movement (Figs. 1D and 1E), and was greater for the anterior region than the posterior region (Table 1; Fig. 2). The overall degree of activation was proportionally greater during voluntary movement than passive movement, being approximately twofold for the anterior region compared to the posterior region. The relative degree of activation between the anterior and the posterior regions during voluntary movement (ratio 24.6:9.6 ⫽ 2.6) was similar to that during passive movement (9.6:4.2 ⫽ 2.3). There were approximately 32 movements performed in each 24-s activation task block for both voluntary and passive movement. Activation was not observed in the contralateral cerebellum during voluntary or passive movement.
672
G.W. Thickbroom et al. / NeuroImage 18 (2003) 670 – 674
Fig. 1. Dual regions of cerebellar activation during voluntary right index finger movement shown for one subject (A) in the transverse plane (two noncontiguous slices, see also (D)), (B) sagittal plane, and (C) coronal plane. Similar dual regions of activation were observed during voluntary (D) and passive (E) finger movement (10 contiguous transverse slices, second subject).
Cerebral cortical activation In all subjects, voluntary movement resulted in activation of the contralateral primary sensorimotor cortex (SM1) and supplementary motor area (SMA), in keeping with the expected pattern of cortical activation during a finger-tapping task. Likewise with passive finger movement, a similar pattern of cortical activation was obtained, but the degree of activation was less than that during the voluntary movement task in both the SM1 and the SMA (Table 1; Fig. 2). Proportionally, the relative pattern of activation was the same for both regions during voluntary and passive movement, with almost double the activation during voluntary movement for both SM1 (ratio 40.8:26.4 ⫽ 1.5) and SMA (10.8:6.2 ⫽ 1.7).
Discussion
stimulation, or passive movement in some studies (Casey et al., 1996; Mima et al., 1999; Seitz and Roland, 1992; Tempel and Perlmutter, 1992), whereas in other studies activation at times equal to that recorded during voluntary movements has been described with passive movement or sensory stimulation of the upper limb (Jueptner et al., 1997; Jueptner and Weiller, 1998; Nitschke et al., 1998). We found cerebellar activation during passive finger movement; however, the overall level of activation was lower than with a comparable voluntary movement. As the sensory input during passive limb movement presumably involves a combination of tactile, joint, and muscle afferent activity, it is Table 1 Number of activated voxels (group mean ⫾ SD) in the cerebellum (anterior and posterior regions) and cerebral cortex (primary sensorimotor cortex and supplementary motor area) for voluntary and passive movement Cerebellum
There has been conflicting evidence for the presence of cerebellar activation with sensory stimulation in previous human functional imaging studies. Cerebellar activation was not found during vibro-tactile stimulation, noxious
Voluntary Passive
Cerebrum
Anterior
Posterior
SM1
SMA
24.6 ⫾ 2.8 9.6 ⫾ 2.3
9.6 ⫾ 2.4 4.2 ⫾ 1.3
40.8 ⫾ 6.4 26.4 ⫾ 4
10.8 ⫾ 2.2 6.2 ⫾ 1.6
G.W. Thickbroom et al. / NeuroImage 18 (2003) 670 – 674
673
Fig. 2. Group mean data comparing the number of activated voxels in the contralateral sensorimotor cortex (SM1), midline supplementary motor area (SMA) and anterior and posterior ipsilateral cerebellar hemispheres during voluntary and passive movement (mean ⫾ standard error).
possible that the cerebellum is involved in discriminating and integrating these multiple channels of sensory information. Studies that have sought to preferentially activate a single sensory modality, such as vibration alone, have generally been unsuccessful in eliciting cerebellar activation (Casey et al., 1996; Mima et al., 1999 Seitz and Roland, 1992; Tempel and Perlmutter, 1992). Thus, integration of sensory information in the cerebellum from a range of receptors may be the basis for the present observations, and while there was no requirement for overt sensory discrimination, a form of sensory pattern processing may have occurred, resulting in a pattern of activation similar to that during a comparable voluntary movement. Although it was not technically possible to monitor EMG activity in the forearm muscles during the fMRI studies, a number of measures were taken to ensure that there was no voluntary contribution by the subjects during the passive movement studies. These included the use of only experienced subjects, and a preliminary instruction and training session. In addition, in a number of subjects EMG recordings from the forearm muscles outside the scanner failed to show motor unit recruitment during the passive finger movement. While motor imagery during the passive protocol is a further consideration, previous studies of motor imagery have indicated that this results in only weak cerebellar activation and is unlikely to account for the degree of activation observed in the present study (Jueptner et al., 1997; Lotze et al., 1999). The possibility was considered that the dual pattern of activation found might be a consequence of the organization of efferent and afferent cerebellar pathways. For example, could the posterior hand representation have a preferential role in processing kinesthetic spinal inputs via the inferior
cerebellar peduncle? Against this is the fact that these fibers are known to terminate widely throughout the cerebellum, overlapping in both the anterior and posterior lobes, as do the cerebral cortical afferents traveling via the middle peduncle (Rothwell, 1994). Consequently, it appears that sensory afferents are widely distributed in the cerebellum, in keeping with the generally uniform neural organization of this organ, and that the dual hand representation cannot be readily explained on purely anatomical grounds. The observation that the dual cerebellar representation in the anesthetized cat is not present in the awake animal (Combs, 1954) is in keeping with activation of diffuse afferent projections from specific body parts throughout the cerebellum in the awake animal. It has been suggested that during anesthesia only the most direct afferent and efferent pathways remain functional, and that this leads to a simplified somatotopy (Rothwell, 1994). A possible explanation for our finding of a dual representation in the awake human with functional imaging is that functional MRI may detect activation of only the most direct afferent and efferent pathways involving the fewest synapses and leading to simultaneous activation of large populations of cells in a way that yields a significant BOLD response. Thus some caution is warranted in ascribing functional significance to the dual representation in the human, and it remains possible that cerebellar processing is in fact undertaken using more widespread and less spatially defined networks. This is also supported by the difficulty in demonstrating a clear somatotopy within the cerebellum comparable to that readily identified in primary motor and sensory cortex (Grodd et al., 2001; Nitschke et al., 1996; Rijntjes et al., 1999). It is possible that the cytoarchitectonic uniformity of the cerebellum and the diffuse patterns of afferent fiber terminations
674
G.W. Thickbroom et al. / NeuroImage 18 (2003) 670 – 674
in the cerebellar cortex mean that the cerebellum is not preferentially structured to process information in a segregated somatotopic manner. Previous studies of voluntary and passive movement have shown that the pattern of activation in cerebral cortex is similar during both of these tasks (Mima et al., 1999; Weiller et al., 1996; Yetkin et al., 1995; Mima et al., 1999). We observed a relative pattern of activation in the sensorimotor cortex and SMA similar to that in the cerebellum during voluntary and passive movement, the degree of activation in all regions being approximately halved during passive movement. This suggests that a similar global network is activated during both of these tasks, and further highlights the close association between kinesthetic sensory and motor processing throughout the brain at both cortical and subcortical levels. The greater degree of overall activation during voluntary compared with passive movement could be explained by increased neural activation or recruitment as a result of sensory and motor integration during voluntary movement. However, although we attempted to do everything possible to make the voluntary and passive movements the same in all respects, there are likely to be differences in the patterns of afferent inputs in the two situations, and we cannot exclude the possibility that the differences we have observed reflect different levels of afferent inputs during voluntary and passive movement. As well, differences in attention during voluntary movement compared to passive movement could affect level of activation in these two conditions. In conclusion, we have shown that it is possible to demonstrate cerebellar activation with passive movement of a single digit using fMRI and that there is a dual pattern of activation as in the case of voluntary movement. While the functional significance of the dual hand representation in the anterior and posterior lobes of the cerebellum remains uncertain, the present findings indicate that both areas are involved in sensory and motor processing.
Acknowledgments The authors are grateful to Dr. S. Ghosh for comments and helpful discussion. Dr. M. Fallon, Dr. S. Davis, and Mr. I. Morris, Department of Radiology, MRI Unit, Sir Charles Gairdner Hospital, and radiographers from the MRI Unit are thanked for support and assistance with the imaging studies. Prof. L.A. Cala and members of the Australian Research Centre for Medical Engineering are thanked for assistance with the passive movement device.
References Casey, K.L., Minoshima, S., Morrow, T.J., Koeppe, R.A., 1996. Comparison of human cerebral activation pattern during cutaneous warmth, heat pain, and deep cold pain. J. Neurophysiol. 76, 571–581. Combs, C.M., 1954. Electro-anatomical study of cerebellar localisation. Stimulation of various afferents. J. Neurosci. 17, 123–143. Grodd, W., Hulsmann, E., Lotze, M., Wildgruber, D., Erb, M., 2001. Sensorimotor mapping of the human cerebellum: fMRI evidence of somatotopic organisation. Human Brain Mapp. 13, 55–73. Jueptner, M., Ottinger, S., Fellows, S.J., Adamschewski, J., Flerich, L., Muller, S.P., Diener, H.C., Thilmann, A.F., Weiller, C., 1997. The relevance of sensory input for the cerebellar control of movements. NeuroImage 5, 41– 48. Jueptner, M., Weiller, C., 1998. A review of differences between basal ganglia and cerebellar control of movements as revealed by functional imaging studies. Brain 121, 1437–1449. Lotze, M., Montoya, P., Erb, M., Hulsmann, E., Flor, H., Klose, U., Birbaumer, N., Grodd, W., 1999. Activation of cortical and cerebellar motor areas during executed and imagined hand movements: an fMRI study. J. Cogn. Neurosci. 11, 491–501. Mima, T., Sadato, N., Yazawa, S., Hanakawa, T., Fukuyama, H., Yonekura, Y., Shibasaki, H., 1999. Brain structures related to active and passive finger movements in man. Brain 122, 1989 –1997. Nitschke, M.F., Hahn, C., Melchert, U.H., Handels, H., Wessel, K., 1998. Activation of the cerebellum by sensory finger stimulation and by finger opposition movements—a functional magnetic resonance imaging study. J. Neuroimaging 8, 127–131. Nitschke, M.F., Kleinschmidt, A., Wessel, K., Frahm, J., 1996. Somatotopic motor representation in the human anterior cerebellum. A highresolution functional MRI study. Brain 119, 1023–1029. Rijntjes, M., Buechel, C., Kiebel, S., Weiller, C., 1999. Multiple somatotopic representations in the human cerebellum. NeuroReport 10, 3653– 3658. Rothwell, J.C., 1994. Control of Human Voluntary Movement. Chapman & Hall, London. Schmahmann, J.D., Doyon, J., McDonald, D., Holmes, C., Lavoie, K., Hurwitz, A.S., Kabani, N., Toga, A., Evans, A., Petrides, M., 1999. Three-dimensional MRI atlas of the human cerebellum in proportional sterotaxic space. NeuroImage 10, 233–260. Seitz, R.J., Roland, P.E., 1992. Vibratory stimulation increases and decreases the regional cerebral blood flow and oxidative metabolism: a positron emission tomography (PET) study. Acta Neurol. Scand. 86, 60 – 67. Snider, R.S., Eldred, E., 1952. Cerebrocerebellar relationships in the monkey. J. Neurophysiol. 15, 27– 40. Tempel, L.W., Perlmutter, J.S., 1992. Vibration-induced regional cerebral blood flow responses in normal ageing. J. Cerebral Blood Flow Metab. 12, 554 –561. Thickbroom, G.W., Byrnes, M.L., Sacco, P., Ghosh, S., Morris, I.T., Mastaglia, F.L., 2000. The role of the supplementary motor area in externally timed movement: the influence of predictability of movement timing. Brain Res. 874, 233–241. Weiller, C., Juptner, M., Fellows, S., Rijntjes, M., Leonhardt, G., Kiebel, S., Muller, S., Diener, H.C., Thilmann, A.F., 1996. Brain representation of active and passive movements. NeuroImage 4, 105–110. Yetkin, F.Z., Muellwe, W.M., Hammeke, T.A., Morris, G.L., Haughton, V.M., 1995. Functional magnetic resonance imaging mapping of sensorimotor cortex with tactile stimulation. Neurosurgery 36, 921– 925.