Attentional activation of the cerebellum independent of motor involvement

Science

Volumn 275, Issue 5308, 1997, Pages 1940-1943

  Allen, Greg ^a,b   Buxton, Richard B ^c   Wong, Eric C ^c   Courchesne, Eric ^b,d  

^a SAN DIEGO STATE UNIVERSITY   (United States)

^b RADY CHILDREN'S HOSPITAL   (United States)

^c UNIVERSITY OF CALIFORNIA   (United States)

^d UNIVERSITY OF CALIFORNIA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

ADULT; ARTICLE; ATTENTION; CEREBELLUM; COGNITION; FEMALE; HUMAN; HUMAN EXPERIMENT; MALE; MOTOR PERFORMANCE; NUCLEAR MAGNETIC RESONANCE IMAGING; PRIORITY JOURNAL;

ADULT; ATTENTION; BRAIN MAPPING; CEREBELLAR CORTEX; CEREBELLUM; COGNITION; FEMALE; HUMANS; MAGNETIC RESONANCE IMAGING; MALE; MOTOR ACTIVITY; PSYCHOMOTOR PERFORMANCE;

EID: 0030995187     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.275.5308.1940     Document Type: Article

References (53)

1. R. W. Williams and K. Herrup, Annu. Rev. Neurosci. 11, 423 (1988).
2. E. Courchesne et al., Behav. Neurosci. 108, 848 (1994).
3. J. D. Schmahmann, Hum. Brain Mapp. 4, 174 (1996); S. Gilman, J. R. Bloedel, R. Lechtenberg, Disorders of the Cerebellum (Davis, Philadelphia, 1981).
4. J. D. Schmahmann, Hum. Brain Mapp. 4, 174 (1996); S. Gilman, J. R. Bloedel, R. Lechtenberg, Disorders of the Cerebellum (Davis, Philadelphia, 1981).
5. L. Crispino and T. H. Bullock, Proc. Natl. Acad. Sci. U.S.A. 81, 2917 (1984).
6. P. P. Newman and H. Reza, J. Physiol. (London) 287, 405 (1979).
7. M. Ito, The Cerebellum and Neural Control (Raven, New York, 1984); R. Llinas and C. Sotelo, The Cerebellum Revisited (Springer-Verlag, New York, 1992); W. T. Thach, H. P. Goodkin, J. G. Keating, Annu. Rev. Neurosci. 15, 403 (1992); V. Braitenberg, D. Heck, F. Sultan, Behav. Brain Sci., in press.
8. M. Ito, The Cerebellum and Neural Control (Raven, New York, 1984); R. Llinas and C. Sotelo, The Cerebellum Revisited (Springer-Verlag, New York, 1992); W. T. Thach, H. P. Goodkin, J. G. Keating, Annu. Rev. Neurosci. 15, 403 (1992); V. Braitenberg, D. Heck, F. Sultan, Behav. Brain Sci., in press.
9. M. Ito, The Cerebellum and Neural Control (Raven, New York, 1984); R. Llinas and C. Sotelo, The Cerebellum Revisited (Springer-Verlag, New York, 1992); W. T. Thach, H. P. Goodkin, J. G. Keating, Annu. Rev. Neurosci. 15, 403 (1992); V. Braitenberg, D. Heck, F. Sultan, Behav. Brain Sci., in press.
10. M. Ito, The Cerebellum and Neural Control (Raven, New York, 1984); R. Llinas and C. Sotelo, The Cerebellum Revisited (Springer-Verlag, New York, 1992); W. T. Thach, H. P. Goodkin, J. G. Keating, Annu. Rev. Neurosci. 15, 403 (1992); V. Braitenberg, D. Heck, F. Sultan, Behav. Brain Sci., in press.
11. J.-H. Gao et al., Science 272, 545 (1996).
12. N. A. Akshoomoff and E. Courchesne, Behav. Neurosci. 106, 731 (1992); J. Cogn. Neurosci. 6, 388 (1994); J. Townsend, E. Courchesne, B. Egaas, Dev. Psychopathol. 8, 563 (1996).
13. N. A. Akshoomoff and E. Courchesne, Behav. Neurosci. 106, 731 (1992); J. Cogn. Neurosci. 6, 388 (1994); J. Townsend, E. Courchesne, B. Egaas, Dev. Psychopathol. 8, 563 (1996).
14. N. A. Akshoomoff and E. Courchesne, Behav. Neurosci. 106, 731 (1992); J. Cogn. Neurosci. 6, 388 (1994); J. Townsend, E. Courchesne, B. Egaas, Dev. Psychopathol. 8, 563 (1996).
15. E. Courchesne, N. A. Akshoomoff, J. Townsend, O. Saitoh, in Perspectives of Event-Related Potentials Research, G. Karmos, M. Molnar, V. Csepe, I. Czigler, J. E. Desmedt, Eds. (Elsevier Science, Amsterdam, 1995), pp. 315-325.
16. T. H. Le and X. Hu, Neuroimage 3, 246 (1996).
17. T. Klingberg, R. Kawashima, P. E. Roland, Eur. J. Neurosci. 8, 1965 (1996).
18. E. Artiges et al., Hum. Brain Mapp. (suppl. 1), 227 (1995); A. Martin, J. V. Haxby, F. M. Lalonde, C. L. Wiggs, L. G. Ungerleider, Science 270, 102 (1995); S. E. Petersen, P. T. Fox, M. I. Posner, M. Mintun, M. E. Raichle, J. Cogn. Neurosci. 1, 153 (1989).
19. E. Artiges et al., Hum. Brain Mapp. (suppl. 1), 227 (1995); A. Martin, J. V. Haxby, F. M. Lalonde, C. L. Wiggs, L. G. Ungerleider, Science 270, 102 (1995); S. E. Petersen, P. T. Fox, M. I. Posner, M. Mintun, M. E. Raichle, J. Cogn. Neurosci. 1, 153 (1989).
20. E. Artiges et al., Hum. Brain Mapp. (suppl. 1), 227 (1995); A. Martin, J. V. Haxby, F. M. Lalonde, C. L. Wiggs, L. G. Ungerleider, Science 270, 102 (1995); S. E. Petersen, P. T. Fox, M. I. Posner, M. Mintun, M. E. Raichle, J. Cogn. Neurosci. 1, 153 (1989).
21. N. C. Andreasen et al., Proc. Natl. Acad. Sci. U.S.A. 92, 5111 (1995); P. M. Grasby et al., Brain 116, 1 (1993).
22. N. C. Andreasen et al., Proc. Natl. Acad. Sci. U.S.A. 92, 5111 (1995); P. M. Grasby et al., Brain 116, 1 (1993).
23. S.-G. Kim, K. Uǧurbil, P. L. Strick, Science 265, 949 (1994).
24. The only exception is a recent paper showing dentate nucleus activation during tactile discrimination (7); all other studies cited above involved an overt motor response.
25. J. M. Bower and J. Kassel, J. Comp. Neurol. 302, 768 (1990).
26. Stimuli (duration 50 ms) appeared randomly at a rate of ∼3 per second (target rate, ∼1 per second; maximum and minimum intertarget intervals, 4630 ms and 240 ms, respectively).
27. Every 40 s over the course of 320 s, a one-word instruction cued the alternation between activation and control conditions. In the Attention tasks, "FORM" cued subjects to respond to each square, "COLOR" cued them to respond to every red shape, and "STOP" cued the control condition. For the Motor and Attention-with-Motor tasks, the movement was the squeezing of a bulb for two subjects and the pressing of a key for the other four. "GO" cued the onset of movement, and "STOP" cued rest.
28. Functional images were acquired at 1.5 T using conventional gradient hardware and a gradient-echo EPI pulse sequence [repetition time (TR) = 2500 ms; echo time (TE) = 40 ms; flip angle = 90°; matrix = 64 × 64; field of view (FOV) = 32 cm; slice thickness = 5 mm; slice gap = 1 mm]. Anatomical images [3D spoiled gradient-echo (SPGR) pulse sequence: TR = 30 ms; TE = 5 ms; flip angle = 45°; matrix = 256 × 192 × 60; FOV = 24 cm; slice thickness = 2 mm] were acquired at the same locations during the same scan session for each subject.
29. To locate significant activations, we used the correlational analysis described in P. A. Bandettini, A. Jesmanowicz, E. C. Wong, J. S. Hyde, Magn. Reson. Med. 30, 161 (1993). The time-course signal data from each voxel were correlated with a model response function, a boxcar wave with sloped sides (duration 7.5 s), to approximate the delay between task onset and maximum signal change. Significantly activated voxels were those that exceeded a threshold r value equivalent to one-tailed P < 0.05 with Bonferroni correction for multiple comparisons, corresponding to r = 0.35 at P < 0.00003.
30. ROIs were based in part on our pilot fMRI studies of attention and motor activation and in part on our previous work showing cerebellar activity during self-paced right-hand movements localized primarily to the ipsilateral, anterior cerebellum [U. Sabatini et al., J. Cereb. Blood Flow Metab. 13, 639 (1993); H. Shibasaki et al., Brain 116, 1387 (1993)], whereas activity during tasks requiring attention to a stimulus guiding right-hand movements extended into the contralateral, posterior cerebellum (28) [M. Blinkenberg, C. Bonde, O. B. Paulson, C. Svarer, I. Law, Hum. Brain Mapp. (suppl. 1), 280 (1995); C. T. Bonde, M. Blinkenberg, I. Law, C. Svarer, O. B. Paulson, ibid., p. 313; J. M. Ellermann et al., NMR Biomed. 7, 63 (1994)]. Together, these findings suggested a functional distinction between cerebellar regions forming the basis for our ROIs with further support from cerebrocerebellar circuitry. For instance, the lateral neocerebellum sends output to the prefrontal cortex [F. A. Middleton and P. L. Strick, Science 266, 458 (1994)]; the right prefrontal cortex is involved in attention tasks of the type we used [D. T. Stuss, T. Shallice, M. P. Alexander, T. W. Picton, Ann. N. Y. Acad. Sci. 769, 191 (1995)]; and cerebellar input to the right cerebrum is from the left cerebellum [M. B. Carpenter, Core Text of Neuroanatomy (Williams & Wilkins, Baltimore, ed. 4, 1991)].
31. ROIs were based in part on our pilot fMRI studies of attention and motor activation and in part on our previous work showing cerebellar activity during self-paced right-hand movements localized primarily to the ipsilateral, anterior cerebellum [U. Sabatini et al., J. Cereb. Blood Flow Metab. 13, 639 (1993); H. Shibasaki et al., Brain 116, 1387 (1993)], whereas activity during tasks requiring attention to a stimulus guiding right-hand movements extended into the contralateral, posterior cerebellum (28) [M. Blinkenberg, C. Bonde, O. B. Paulson, C. Svarer, I. Law, Hum. Brain Mapp. (suppl. 1), 280 (1995); C. T. Bonde, M. Blinkenberg, I. Law, C. Svarer, O. B. Paulson, ibid., p. 313; J. M. Ellermann et al., NMR Biomed. 7, 63 (1994)]. Together, these findings suggested a functional distinction between cerebellar regions forming the basis for our ROIs with further support from cerebrocerebellar circuitry. For instance, the lateral neocerebellum sends output to the prefrontal cortex [F. A. Middleton and P. L. Strick, Science 266, 458 (1994)]; the right prefrontal cortex is involved in attention tasks of the type we used [D. T. Stuss, T. Shallice, M. P. Alexander, T. W. Picton, Ann. N. Y. Acad. Sci. 769, 191 (1995)]; and cerebellar input to the right cerebrum is from the left cerebellum [M. B. Carpenter, Core Text of Neuroanatomy (Williams & Wilkins, Baltimore, ed. 4, 1991)].
32. ROIs were based in part on our pilot fMRI studies of attention and motor activation and in part on our previous work showing cerebellar activity during self-paced right-hand movements localized primarily to the ipsilateral, anterior cerebellum [U. Sabatini et al., J. Cereb. Blood Flow Metab. 13, 639 (1993); H. Shibasaki et al., Brain 116, 1387 (1993)], whereas activity during tasks requiring attention to a stimulus guiding right-hand movements extended into the contralateral, posterior cerebellum (28) [M. Blinkenberg, C. Bonde, O. B. Paulson, C. Svarer, I. Law, Hum. Brain Mapp. (suppl. 1), 280 (1995); C. T. Bonde, M. Blinkenberg, I. Law, C. Svarer, O. B. Paulson, ibid., p. 313; J. M. Ellermann et al., NMR Biomed. 7, 63 (1994)]. Together, these findings suggested a functional distinction between cerebellar regions forming the basis for our ROIs with further support from cerebrocerebellar circuitry. For instance, the lateral neocerebellum sends output to the prefrontal cortex [F. A. Middleton and P. L. Strick, Science 266, 458 (1994)]; the right prefrontal cortex is involved in attention tasks of the type we used [D. T. Stuss, T. Shallice, M. P. Alexander, T. W. Picton, Ann. N. Y. Acad. Sci. 769, 191 (1995)]; and cerebellar input to the right cerebrum is from the left cerebellum [M. B. Carpenter, Core Text of Neuroanatomy (Williams & Wilkins, Baltimore, ed. 4, 1991)].
33. ROIs were based in part on our pilot fMRI studies of attention and motor activation and in part on our previous work showing cerebellar activity during self-paced right-hand movements localized primarily to the ipsilateral, anterior cerebellum [U. Sabatini et al., J. Cereb. Blood Flow Metab. 13, 639 (1993); H. Shibasaki et al., Brain 116, 1387 (1993)], whereas activity during tasks requiring attention to a stimulus guiding right-hand movements extended into the contralateral, posterior cerebellum (28) [M. Blinkenberg, C. Bonde, O. B. Paulson, C. Svarer, I. Law, Hum. Brain Mapp. (suppl. 1), 280 (1995); C. T. Bonde, M. Blinkenberg, I. Law, C. Svarer, O. B. Paulson, ibid., p. 313; J. M. Ellermann et al., NMR Biomed. 7, 63 (1994)]. Together, these findings suggested a functional distinction between cerebellar regions forming the basis for our ROIs with further support from cerebrocerebellar circuitry. For instance, the lateral neocerebellum sends output to the prefrontal cortex [F. A. Middleton and P. L. Strick, Science 266, 458 (1994)]; the right prefrontal cortex is involved in attention tasks of the type we used [D. T. Stuss, T. Shallice, M. P. Alexander, T. W. Picton, Ann. N. Y. Acad. Sci. 769, 191 (1995)]; and cerebellar input to the right cerebrum is from the left cerebellum [M. B. Carpenter, Core Text of Neuroanatomy (Williams & Wilkins, Baltimore, ed. 4, 1991)].
34. ROIs were based in part on our pilot fMRI studies of attention and motor activation and in part on our previous work showing cerebellar activity during self-paced right-hand movements localized primarily to the ipsilateral, anterior cerebellum [U. Sabatini et al., J. Cereb. Blood Flow Metab. 13, 639 (1993); H. Shibasaki et al., Brain 116, 1387 (1993)], whereas activity during tasks requiring attention to a stimulus guiding right-hand movements extended into the contralateral, posterior cerebellum (28) [M. Blinkenberg, C. Bonde, O. B. Paulson, C. Svarer, I. Law, Hum. Brain Mapp. (suppl. 1), 280 (1995); C. T. Bonde, M. Blinkenberg, I. Law, C. Svarer, O. B. Paulson, ibid., p. 313; J. M. Ellermann et al., NMR Biomed. 7, 63 (1994)]. Together, these findings suggested a functional distinction between cerebellar regions forming the basis for our ROIs with further support from cerebrocerebellar circuitry. For instance, the lateral neocerebellum sends output to the prefrontal cortex [F. A. Middleton and P. L. Strick, Science 266, 458 (1994)]; the right prefrontal cortex is involved in attention tasks of the type we used [D. T. Stuss, T. Shallice, M. P. Alexander, T. W. Picton, Ann. N. Y. Acad. Sci. 769, 191 (1995)]; and cerebellar input to the right cerebrum is from the left cerebellum [M. B. Carpenter, Core Text of Neuroanatomy (Williams & Wilkins, Baltimore, ed. 4, 1991)].
35. ROIs were based in part on our pilot fMRI studies of attention and motor activation and in part on our previous work showing cerebellar activity during self-paced right-hand movements localized primarily to the ipsilateral, anterior cerebellum [U. Sabatini et al., J. Cereb. Blood Flow Metab. 13, 639 (1993); H. Shibasaki et al., Brain 116, 1387 (1993)], whereas activity during tasks requiring attention to a stimulus guiding right-hand movements extended into the contralateral, posterior cerebellum (28) [M. Blinkenberg, C. Bonde, O. B. Paulson, C. Svarer, I. Law, Hum. Brain Mapp. (suppl. 1), 280 (1995); C. T. Bonde, M. Blinkenberg, I. Law, C. Svarer, O. B. Paulson, ibid., p. 313; J. M. Ellermann et al., NMR Biomed. 7, 63 (1994)]. Together, these findings suggested a functional distinction between cerebellar regions forming the basis for our ROIs with further support from cerebrocerebellar circuitry. For instance, the lateral neocerebellum sends output to the prefrontal cortex [F. A. Middleton and P. L. Strick, Science 266, 458 (1994)]; the right prefrontal cortex is involved in attention tasks of the type we used [D. T. Stuss, T. Shallice, M. P. Alexander, T. W. Picton, Ann. N. Y. Acad. Sci. 769, 191 (1995)]; and cerebellar input to the right cerebrum is from the left cerebellum [M. B. Carpenter, Core Text of Neuroanatomy (Williams & Wilkins, Baltimore, ed. 4, 1991)].
36. ROIs were based in part on our pilot fMRI studies of attention and motor activation and in part on our previous work showing cerebellar activity during self-paced right-hand movements localized primarily to the ipsilateral, anterior cerebellum [U. Sabatini et al., J. Cereb. Blood Flow Metab. 13, 639 (1993); H. Shibasaki et al., Brain 116, 1387 (1993)], whereas activity during tasks requiring attention to a stimulus guiding right-hand movements extended into the contralateral, posterior cerebellum (28) [M. Blinkenberg, C. Bonde, O. B. Paulson, C. Svarer, I. Law, Hum. Brain Mapp. (suppl. 1), 280 (1995); C. T. Bonde, M. Blinkenberg, I. Law, C. Svarer, O. B. Paulson, ibid., p. 313; J. M. Ellermann et al., NMR Biomed. 7, 63 (1994)]. Together, these findings suggested a functional distinction between cerebellar regions forming the basis for our ROIs with further support from cerebrocerebellar circuitry. For instance, the lateral neocerebellum sends output to the prefrontal cortex [F. A. Middleton and P. L. Strick, Science 266, 458 (1994)]; the right prefrontal cortex is involved in attention tasks of the type we used [D. T. Stuss, T. Shallice, M. P. Alexander, T. W. Picton, Ann. N. Y. Acad. Sci. 769, 191 (1995)]; and cerebellar input to the right cerebrum is from the left cerebellum [M. B. Carpenter, Core Text of Neuroanatomy (Williams & Wilkins, Baltimore, ed. 4, 1991)].
37. ROIs were based in part on our pilot fMRI studies of attention and motor activation and in part on our previous work showing cerebellar activity during self-paced right-hand movements localized primarily to the ipsilateral, anterior cerebellum [U. Sabatini et al., J. Cereb. Blood Flow Metab. 13, 639 (1993); H. Shibasaki et al., Brain 116, 1387 (1993)], whereas activity during tasks requiring attention to a stimulus guiding right-hand movements extended into the contralateral, posterior cerebellum (28) [M. Blinkenberg, C. Bonde, O. B. Paulson, C. Svarer, I. Law, Hum. Brain Mapp. (suppl. 1), 280 (1995); C. T. Bonde, M. Blinkenberg, I. Law, C. Svarer, O. B. Paulson, ibid., p. 313; J. M. Ellermann et al., NMR Biomed. 7, 63 (1994)]. Together, these findings suggested a functional distinction between cerebellar regions forming the basis for our ROIs with further support from cerebrocerebellar circuitry. For instance, the lateral neocerebellum sends output to the prefrontal cortex [F. A. Middleton and P. L. Strick, Science 266, 458 (1994)]; the right prefrontal cortex is involved in attention tasks of the type we used [D. T. Stuss, T. Shallice, M. P. Alexander, T. W. Picton, Ann. N. Y. Acad. Sci. 769, 191 (1995)]; and cerebellar input to the right cerebrum is from the left cerebellum [M. B. Carpenter, Core Text of Neuroanatomy (Williams & Wilkins, Baltimore, ed. 4, 1991)].
38. ROI drawing was guided by a human cerebellar atlas [G. A. Press, J. W. Murakami, E. Courchesne, M. Grafe, J. R. Hesselink, Am. J. Neuroradiol. 11, 41 (1990)]. The Motor ROI was drawn from the surface location of the right primary fissure (pf) to the center of the band of white matter separating the anterior vermis (AVe) from the posterior vermis. From there, a line was drawn to the apex of the AVe. This ROI was completed by a line drawn along the surface of the cerebellum, back to the right pf. The Attention ROI was drawn from the surface location of the left pf to the center of the same white-matter band. A second line was drawn from this point to the surface location of the left horizontal fissure (hf). A line drawn along the surface of the cerebellum back to the left pf completed this ROI.
39. Because all stimuli were presented at a single spatial location in the center of foveal vision, eye movement activation was not predicted to occur. Moreover, previous work would predict that if eye movements had occurred, the resulting activation would occur in the cerebellar vermis [L. Petit et al., J. Neurosci. 16, 3714 (1996)], a region not activated during the Attention task. All areas that were active during the Attention task were also active during the Attention-with-Motor task, indicating that those behavioral requirements unique to the Attention task-namely, silent counting and encoding the number of targets - did not add to the results. Silent counting activation has been reported in the inferior cerebellum [E. Ryding, J. Decety, H. Sjoholm, G. Stenberg, D. H. Ingvar, Brain Res. Cogn. Brain Res. 1, 94 (1993)] and in a "midline cerebellar region" [J. A. Fiez et al., J. Neurosci. 16, 808 (1996)]. In our study, the focus of attention activity was in the left cerebellar hemisphere, yet the right cerebellar hemisphere is consistently active during verbal tasks (12). Thus, nonverbal visual attention activated a side and region inconsistent with predicted silent counting effects. Still, to investigate whether silent counting might have contributed to these results, we instructed four subjects to silently count from 1 to 10 repeatedly in the absence of any visual stimuli. Examination of activation during this task revealed no cerebellar activation within the Attention VOI. Working memory activation of the cerebellum has also been reported (11), and the requirement to encode the number of targets during the Attention task placed minor demands on working memory. However, encoding the number targets was not required by the Attention-with-Motor task, and there were no regions of cerebellar activation unique to the Attention task as compared with the Attention-with-Motor task. Thus, like silent counting, working memory did not contribute to the activation effects observed during the Attention task.
40. Because all stimuli were presented at a single spatial location in the center of foveal vision, eye movement activation was not predicted to occur. Moreover, previous work would predict that if eye movements had occurred, the resulting activation would occur in the cerebellar vermis [L. Petit et al., J. Neurosci. 16, 3714 (1996)], a region not activated during the Attention task. All areas that were active during the Attention task were also active during the Attention-with-Motor task, indicating that those behavioral requirements unique to the Attention task-namely, silent counting and encoding the number of targets - did not add to the results. Silent counting activation has been reported in the inferior cerebellum [E. Ryding, J. Decety, H. Sjoholm, G. Stenberg, D. H. Ingvar, Brain Res. Cogn. Brain Res. 1, 94 (1993)] and in a "midline cerebellar region" [J. A. Fiez et al., J. Neurosci. 16, 808 (1996)]. In our study, the focus of attention activity was in the left cerebellar hemisphere, yet the right cerebellar hemisphere is consistently active during verbal tasks (12). Thus, nonverbal visual attention activated a side and region inconsistent with predicted silent counting effects. Still, to investigate whether silent counting might have contributed to these results, we instructed four subjects to silently count from 1 to 10 repeatedly in the absence of any visual stimuli. Examination of activation during this task revealed no cerebellar activation within the Attention VOI. Working memory activation of the cerebellum has also been reported (11), and the requirement to encode the number of targets during the Attention task placed minor demands on working memory. However, encoding the number targets was not required by the Attention-with-Motor task, and there were no regions of cerebellar activation unique to the Attention task as compared with the Attention-with-Motor task. Thus, like silent counting, working memory did not contribute to the activation effects observed during the Attention task.
41. Because all stimuli were presented at a single spatial location in the center of foveal vision, eye movement activation was not predicted to occur. Moreover, previous work would predict that if eye movements had occurred, the resulting activation would occur in the cerebellar vermis [L. Petit et al., J. Neurosci. 16, 3714 (1996)], a region not activated during the Attention task. All areas that were active during the Attention task were also active during the Attention-with-Motor task, indicating that those behavioral requirements unique to the Attention task-namely, silent counting and encoding the number of targets - did not add to the results. Silent counting activation has been reported in the inferior cerebellum [E. Ryding, J. Decety, H. Sjoholm, G. Stenberg, D. H. Ingvar, Brain Res. Cogn. Brain Res. 1, 94 (1993)] and in a "midline cerebellar region" [J. A. Fiez et al., J. Neurosci. 16, 808 (1996)]. In our study, the focus of attention activity was in the left cerebellar hemisphere, yet the right cerebellar hemisphere is consistently active during verbal tasks (12). Thus, nonverbal visual attention activated a side and region inconsistent with predicted silent counting effects. Still, to investigate whether silent counting might have contributed to these results, we instructed four subjects to silently count from 1 to 10 repeatedly in the absence of any visual stimuli. Examination of activation during this task revealed no cerebellar activation within the Attention VOI. Working memory activation of the cerebellum has also been reported (11), and the requirement to encode the number of targets during the Attention task placed minor demands on working memory. However, encoding the number targets was not required by the Attention-with-Motor task, and there were no regions of cerebellar activation unique to the Attention task as compared with the Attention-with-Motor task. Thus, like silent counting, working memory did not contribute to the activation effects observed during the Attention task.
42. In rats, when cerebellar stimulation occurs in advance of a sensory stimulus, neural responsiveness to that stimulus is altered at the brainstem, thalamic, hippocampal, and cortical levels (4, 5), and neural signal-to-noise enhancement can result (4); such effects are independent of the engagement of motor systems. For instance, when background luminance reduces to noise levels the colliculus response to a flash, stimulation of vermis lobules VI-VII causes the colliculus response to that flash to emerge above noise if stimulation occurs in advance of the visual stimulus (4).
43. M. G. Paulin, Brain Behav. Evol. 41, 39 (1993).
44. J. C. Eccles, M. Ito, J. Szentagothai, The Cerebellum as a Neuronal Machine (Springer-Verlag, Berlin, 1967).
45. D. Flament, J. M. Ellermann, S.-G. Kim, K. Uǧurbil, T. J. Ebner, Hum. Brain Mapp. 4, 210 (1996); M. E. Raichle et al., Cereb. Cortex 4, 8 (1994).
46. D. Flament, J. M. Ellermann, S.-G. Kim, K. Uǧurbil, T. J. Ebner, Hum. Brain Mapp. 4, 210 (1996); M. E. Raichle et al., Cereb. Cortex 4, 8 (1994).
47. S. L. Rao et al., Hum. Brain Mapp. (suppl. 1), 412 (1995).
48. When sensory information is anticipated, attention is quickly and accurately redirected toward the predicted source of information. On the basis of neurobehavioral and neurophysiological evidence in patients with cerebellar lesions, it has been hypothesized that the cerebellum, through its connections with attention systems (2), influences the speed and accuracy of such attention changes (2, 9).
49. The cerebellum accomplishes this anticipatory function by encoding ("learning") sequences of multidimensional information about external and internal events. A large body of evidence shows that the cerebellum may be involved in such learning [J. L. Raymond, S. G. Lisberger, M. D. Mauk, Science 272, 1126 (1996)]. Whenever an analogous sequence begins to unfold, the cerebellum predicts what is about to happen, reads out the rest of the sequence, and triggers changes in the neural responsiveness of systems expected to be needed in upcoming moments (2, 9).
50. Anticipation involves predicting the internal conditions needed for a particular motor or mental operation and setting those conditions in preparation for that operation. Complete knowledge of upcoming events is not necessary; simple exposure to aspects of a stimulus that may soon arrive will trigger anticipatory responding of the cerebellum. The anticipatory response is neither a sensory nor a motor activity, but rather a general response that prepares whichever neural systems may be necessary in upcoming moments. An example may be changes in the vestibulo-ocular reflex (VOR) in anticipation of changes in vergence observed in the monkey [L. H. Snyder, D. M. Lawrence, W. M. King, Vision Res. 32, 569 (1992)]. A model of how the cerebellum might mediate such anticipatory modulation of the VOR has been proposed [O. Coenen and T. J. Sejnowski, in Advances in Neural Information Processing 8, D. Touretzky, M. Mozer, M. Hasselmo, Eds. (MIT Press, Cambridge, MA, 1996), pp. 89-95].
51. Anticipation involves predicting the internal conditions needed for a particular motor or mental operation and setting those conditions in preparation for that operation. Complete knowledge of upcoming events is not necessary; simple exposure to aspects of a stimulus that may soon arrive will trigger anticipatory responding of the cerebellum. The anticipatory response is neither a sensory nor a motor activity, but rather a general response that prepares whichever neural systems may be necessary in upcoming moments. An example may be changes in the vestibulo-ocular reflex (VOR) in anticipation of changes in vergence observed in the monkey [L. H. Snyder, D. M. Lawrence, W. M. King, Vision Res. 32, 569 (1992)]. A model of how the cerebellum might mediate such anticipatory modulation of the VOR has been proposed [O. Coenen and T. J. Sejnowski, in Advances in Neural Information Processing 8, D. Touretzky, M. Mozer, M. Hasselmo, Eds. (MIT Press, Cambridge, MA, 1996), pp. 89-95].
52. To create functional maps, we interpolated the correlation coefficient images to match the resolution of anatomical images and registered them to the anatomical images to reduce warping. Next, through rotation, translation, and scaling, each subject's cerebellum was transformed to a standard anatomical space by normalizing to a single subject chosen as the standard. All activated voxels were then superimposed across subjects for each task and slice.
53. We thank S. A. Hilyard, T. J. Sejnowski, and J. Townsend for helpful comments; L. Frank, D. Blasttier, and M. Belmonte for technical assistance; and the Santarsiero family for a generous donation. Supported by National Institute of Mental Health grant RO1-MH36840 (E.C.), a San Diego Children's Hospital Research Center seed grant (GA), and a McDonnell-Pew Graduate Fellowship in Cognitive Neuroscience (G.A).