- Email: [email protected]
Blood flow velocity changes in the middle cerebral artery induced by processing of hierarchical visual stimuli
Neuropsychologia,Vol.
~ )
Pergamon
PII: S0028 3932(97)00038 9
35. No. 8, pp. 1181 1184, 1997 ~', 1997 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0028 3932 9 7 $17.00 + 0.00
Note Blood flow velocity changes in the middle cerebral artery induced by processing of hierarchical visual stimuli CORNELIA SCHNITTGER,* SONKE JOHANNES,* ANOUSCHEH ARNAVAZ* and THOMAS F. MUNTEt *Department of Neurology, Medizinische Hochschule, Hannover, Germany; tDepartment of Cognitive Science, University of California at San Diego, La Jolla, CA 92093-0515, U.S.A. (Received 31 July 1996; accepted 22 February 1997)
Abstraet--Transcranial Doppler ultrasound was used to measure mean flow velocities in both middle cerebral arteries while 16 young subjects performed a visual task involving the processing of hierarchically structured stimuli. Specifically, large (global) letters composed of smaller (local) letters were presented, with the subjects' task being to attend either to the local or to the global level and press a button whenever a target on the designated level occurred. Each run was comprised of a 35-sec period of passive stimulation, followed by 65 sec of active task. A highly significant increase of blood flow was detected upon initiation of the active task, which was clearly present after ca. 4 sec, The flow velocity reached a maximum after 20 sec and remained stable for the remainder of the active condition. No hemispheric differences with respect to global or local conditions were observed, i : 1997 Elsevier Science Ltd Key Words: blood flow velocity; global/local; transcranial Doppler; hemisphere.
for global/local processing, while the evidence for a role of interhemispheric integration for global interference was contradictory [23, 26]. On the other hand, lesion studies in human patients suggested that the right hemisphere subserves the processing of global and the left hemisphere subserves the processing of local information [13, 14, 20, 22]. The temporoparietal junction area has been implicated as being essential for the processing of global and local level information. In several event-related brain potential studies our group has found evidence for parallel processing of global and local level information ill partially lateralized brain systems [6], with a more recent study revealing larger effects of" target detection at left hemispheric recording sites [8]. The present study sought to further investigate the possible lateralization of global/local processing by assessing changes in blood flow of the middle cerebral arteries with transcranial Doppler ultrasound (TCD), while subjects selectively attended to global and local levels of hierarchical visual stimuli. TCD is a non-invasive method with an excellent time resolution with respect to other techniques for the assessment of cerebral blood flow, such as positron emission tomography and SPECT, permitting the monitoring of blood flow velocity changes over a few seconds bilaterally. Assuming that the diameter of the large basal cerebral arteries remains stable, any change in velocity will reflect a change in flow [7, 10]. Several TCD studies have already suggested that diameter changes in the small resist-
Introduction Visual stimuli can be conceptualized as being comprised of information on different levels (e.g., a face is made up from features like eyes, nose, mouth, etc.), which have been termed global and local levels. These levels are thus defined by their relative place in a hierarchy [9, 11, 12, 16-18]. There have been a number of studies suggesting a hemispheric specialization of the processing of global and local level information. For example, when hierarchically constructed stimuli were presented to the visual half-fields of normal subjects a left visual field reaction time advantage was found for the identification of larger letters. while smaller letters were processed faster in the right visual field [25]. Other authors failed to find this left visual half-field advantage in reaction time for identification of larger letters in normal subjects [16] or did not find any visual field asymmetries corresponding to the identification of global or local letters [1, 2]. Studies with commisurotomized subjects using the same technique did not find a specialization of the two hemispheres
t Address for correspondence: Department of Cognitive Science, University of California at San Diego, La Jolla, CA 920930515, U.S.A.; e-mail: tmuente(a~cogsci.ucsd.edu; tel.: + 1-619534-2440; fax: + 1-619-534-1128. 1181
1182
Note
ance vessels caused by metabolic changes due to cerebral activation resulted in velocity changes in the corresponding larger arteries (e.g., [3-5, 15, 19, 23]).
Methods
Subjects Sixteen right-handed healthy volunteers (mean age 25 years, range 20-33 years, eight women) participated in the experiment. None had a family history of left-handedness. All subjects had normal or corrected-to-normal vision.
errors. During the first 35 sec of a run they passively viewed the stimuli. At this time the letter 'H' or 'S' appeared in the middle of the screen and indicated which target letter had to be responded to for the remaining 65 sec of the run (e.g., 'global S' or 'local H'). Forty runs of each attention level were administered during two experimental sessions. Subjects had to respond to target stimuli by pressing a button held in the right or left hand (hands changed every 10 runs). They sat with their head resting against the back of a reclining chair. Pulse rate was measured continuously and blood pressure was obtained at the beginning and the end of each block, consisting of 10 runs, to ensure that significant variations of these parameters did not occur. Response speed and accuracy were assessed as well. All behavioral and TCD measures were analysed with repeated measures analysis of variance (ANOVA).
Equipment Results A commercially available Doppler unit (Multi Dop X 4, DWL Elektronische Systeme GmbH, Sipplingen, Germany) was employed allowing on-line as well as off-line calculation of mean flow velocity (Vmo,n). During the recording session, markers were set automatically at the beginning and the end of the stimulation period to identify relevant episodes of cognitive stimulation. The two Doppler probes were attached to the right and left temporal ultrasound windows above the zygomatic arch by using a special probe holder. Both middle cerebral arteries were insonated at a depth of 45-55 mm depending on the best signal-to-noise ratio.
Reaction times to targets on the global and local levels were 407msec (S.D. 26msec) and 392msec (S.D. 32msec) ( P < 0.003), respectively. The hit rate was 93.1% (S.D. 6.1%) for targets on the global level and 92.0% (S.D. 6.2%) on the local level (n.s.). No significant changes of heart rate or blood pressure were observed during the task. Figure 2A and B illustrates the attention-related changes in (A)
115
Stimuli and procedure The stimuli consisted of the white letters 'H', 'E', 'S' and 'A' presented in the middle of a black video screen. Global letters were constructed from the appropriate placement of local letters, for example a large H made from small Es in a 5 x 5 matrix (see Fig. 1). Global letters were 7.4 times as tall as local letters. The height-to-width ratio of both global and local letters was 3:2. At the viewing distance of 70cm, the global letters subtended 3.7 c'of visual angle in height. The stimuli were flashed in a rapid sequence with a stimulus duration of 100 msec, and an interstimulus interval varying randomly between 800 and 1200 msec. Before each run, subjects were informed on the level (global/local) to be attended during the run. They were also instructed to respond as quickly as possible while minimizing
110
_o o
105
100
95 0
20
40
60
80
(B)
[
115
I ~
Local right MCA [
/
I
110t
E E E E EEEEE E E E E Fig. 1. Example of the stimuli used in the study. A large letter (global level) is made up from smaller letters (local level). The subjects had to attend either to the global or the local level.
100
Time (see)
[
I 0
20
40
60
80
100
Time (sec) Fig. 2. Blood flow velocity in the left and right middle cerebral arteries (passive stimulation baseline = 100%). (A) Mean blood flow response (16 subjects) for the runs with attention to the global level. (B) The same for attention to the local level. In both cases a steep increase for the first few seconds is seen, followed by a plateau phase. Termination of the task causes a brief overshoot of flow, followed by a return to baseline levels.
Note mean flow velocities. A marked and immediate increase of blood flow velocity was observed at the beginning of the active condition. Compared to the level of flow velocity during passive stimulation, the flow increase was 7.1-9.8% for the right and 6.5 8.4% for the left middle cerebral artery for the individual subjects. The change in flow velocity became apparent within the first 4sec of active viewing, reached a steady state after approximately 20 sec and remained relatively constant for the duration of the task. There is a small overshoot preceding the decrease at the end of active stimulation, which is present in most silently performed tasks, and the maximum decrease of flow velocity is reached approximately 16 sec later, followed by an increase to passive stimulation steady state. The statistics were performed in several steps.
E2l~,cts during the.first 20 sec First, an ANOVA including the 10 2-sec periods of the first 20 sec of the active tasks as levels of a factor showed a highly significant increase of blood flow over these periods [F(9,135)= 12.44, P<0.0001], while no main effects were seen for the factors level [global vs local, F(1,15)=1.17, n.s.] and artery [left vs right, F(1,15)=1.87, n.s.]. Neither the levelx artery nor the level x artery x period interactions approached significance. To assess the onset time of the blood flow response in this task, pairwise comparison of the 2-sec intervals with a 2-sec period during the passive stimulation baseline were done, revealing significant effects for all 2-sec periods except for the 0-2 sec interval [all F(I,15)>7.1, all P<0.03]. As the physical stimulation was identical during the passive stimulation baseline and the active stimulation phase, the blood flow response attributable purely to the task occurs within the first 4 sec.
EIl~'cts durhtg the entire recording epoch An ANOVA using the 10 successive 10-sec time intervals of the entire experimental run (see Fig. 2) as levels of a factor revealed a highly significant effect for this factor [F(9,135) = 31.34, P < 0.0001 ]. The factor level approached significance, with the overall flow velocity in the local condition being slightly faster than in the global condition [F(I, 15)= 2.44, P = 0.13]. Neither the factor artery nor any of the interaction effects reached significance. The pairwise comparisons of the single 10-sec periods with the 50-60-sec period (i.e. right before the end of the active task) revealed no significant differences for the 10-20, 20 30, 30-40 and 40 50-sec periods (all F<2.5, n.s.). However, there were significant differences between the 50 60sec and the 0-10-sec intervals [F(1,15)=9.86, P<0.01] and the 60 70-sec interval [F(1,15)=7.18, P<0.03]. Similarly, there was a difference between the 50-60-sec interval and the last two intervals (80 90 and 90-100 sec) during the passive rest [both F(1,15)> 16. P<0.005].
Discussion The current study had two major goals: (i) to demonstrate the feasibility of the TCD method for the assessment of the time course of changes in blood flow related to visual attention, and (ii) to show differential blood flow changes in the two middle cerebral arteries as a function of attention to either global or local features of a visual stimulus. Regarding the first goal, clear evidence of a rapid increase in cerebral blood flow was provided when subjects changed from passive to active viewing of the visual stimuli. In fact, a sig-
1183 nificant increase of blood flow velocity was found within the first 4sec of the active condition. The blood flow reached a stable plateau at about 20 sec for the rest of the run. At the termination of the active viewing period, there was an overshoot of flow velocity with respect to this plateau. This is in line with previous investigations using TCD in cognitive tasks that showed a similar overshoot of blood flow velocity alter termination of the task (e.g., [24]). With respect to the second point of interest, no evidence was found for a differential lateralized change in blood flow as a function of attending to either global or local levels of a stimulus. Several considerations might help to understand this result. The first regards the sensitivity of the Doppler technique. It might be argued that any asymmetries in blood flow change were too small to be detected with TCD. The middle cerebral artery is the main vessel supplying the brain and it might well be that a circumscribed asymmetric blood flow change in a branch of this artery went by undetected in TCD. It is of interest in this respect to review the data of Hartje et al. [5], who assessed blood flow in the middle cerebral artery with TCD during three verbal and three non-verbal tasks. They were able to show a significant asymmetry towards the left middle cerebral artery in all three verbal tasks, while only one non-verbal task (picture matching) led to a significant asymmetry towards the right. The two other non-verbal tasks (cube comparison, figure assembly) that were thought to tax the right hemisphere more than the left on the basis of previous reports did not show a significant difference. Thus, from the data of Hartje et al. [5], it can be concluded that the TCD technique is, in principle, capable of detection of asymmetries in blood flow of the middle cerebral artery. However, in some instances in which asymmetries are expected, these are not observed with TCD. This might, however, be partly due to the fact that both hemispheres contribute to a particular task. This latter line of reasoning might also apply to the current pattern of data. In fact, there is evidence that the processing of such stimuli depends, to a large degree, on the conjoined activity of both hemispheres. For example, Robertson and Lamb [21] have presented a model that envisions the coordinated efforts of four subsystems in the processing of global and local information: (i) a global processor located in the right posterior superior temporoparietal cortex, (ii) a local processor associated with the homologous left hemisphere region, (iii) a bilateral parietal lobe system controlling the distribution of the attentional resources to the global and local levels, and (iv) posterior temporal temporal pathways used to communicate between the subsystems and to fine tune them. One can inlerpret the recent data of Weekes et al. [26] in a similar vein; they found no evidence for a specialization of the hemispheres of global/local information in two commissurotomized subjects. It might well be that the data obtained from patients with lesions in the temporoparietal region, which suggested a differential capability of the two hemispheres for the processing of either global or local information, might reflect the selective breakdown of a bihemispheric system that, when intact, utilizes both hemispheres to a roughly equal extent. Thus, the current TCD data might adequately reflect the state of hemispheric involvement in the global local task after all, namely a largely bilaterally distributed multicomponent processing system. To summarize, the TCD technique appears to be a promising tool in the investigation of cerebral blood flow during cognition. Probably more important than aspects of lateralization, the TCD technique appears capable of capturing the temporal dynamics of blood flow during a cognitive task. An important step forward would be the demonstration of behavioral changes as a function of blood flow changes. Experiments to that effect are currently under way in our laboratory. Acknowled#ements- We thank J. Kilian tbr excellent technical
1184 support, and Berdieke Wieringa, Karin Weissenborn and Mike Matzke for helpful discussions. T.F.M. was supported by a grant from the Hermann and Lilly Schilling Foundation (TS 013/177/96).
References
1. Abivasatos, B. and Wilding, J. Hemispheric differences in matching Stroop-type letter stimuli. Cortex 18, 5-22, 1982. 2. Boles, D. B. Global versus local processing: is there a hemispheric dichotomy? Neuropsychologia 22, 445 455, 1984. 3. Droste, D. W., Harders, A. G. and Rastogi, E. Two transcranial Doppler studies on blood flow velocity in both middle cerebral arteries during rest and the performance of cognitive tasks. Neuropsychologia 27, 1221-1230, 1989. 4. Harders, A. G., Laborde, G., Droste, D. W. and Rastogi, E. Brain activity and blood flow velocity changes: a transcranial Doppler study. International Journal of Neuroscience 47, 91-102, 1988. 5. Hartje, W., Ringelstein, E. B., Kistinger, B., Fabianek, D. and Willmes, K. Transcranial Doppler ultrasonic assessment of middle cerebral artery blood flow velocity changes during verbal and visuospatial cognitive tasks. Neuropsychologia 32, 1443-1452, 1994, 6. Heinze, H. J. and Mtinte, T. F. Electrophysiological correlates of hierachical stimulus processing: dissociation between onset and later stages of global and local target processing. Neuropsychologia 31, 841 852, 1993. 7. Huber, P. and Handa, J. Effects of contrast material, hypercapnia, hyperventilation, hypertonic glucose and papaverine on the diameter of the cerebral arteries. Investigative Radiology 2, 17 32, 1967. 8. Johannes, S., Wieringa, B. M., Matzke, M. and Mante, T. F. Hierarchical visual stimuli: electrophysiological evidence for separate left hemispheric global and local processing mechanisms. Neuroscience Letters 210, l 11-114, 1996. 9. Kinchla, R. A. and Wolfe, J. M. The order of visual processing: "Top-down", "Bottom-up", or "Middleout". Perception and Psychophysics 25, 225 231, 1979. 10. Kontos, H. A. Validity of cerebral arterial blood flow calculations from velocity measurements. Stroke 20, l 3, 1989. 11. Lamb, M. R. and Robertson, L. C. The processing of hierarchical stimuli: effects of retinal locus, locational uncertainty, and stimulus identity. Perception and Psychophysics 44, 172 181, 1988. 12. Lamb, M. R. and Robertson, L. C. Do response
Note time advantage and interference reflect the order of processing of global- and local-level information? Perception and Psychophysics 46, 254-258, 1989. 13. Lamb, M. R., Robertson, L. C. and Knight, R. T. Attention and interference in the processing of global and local information: effects of unilateral temporal parietal junction lesions. Neuropsychologia 27, 47 l 483, 1989. 14. Lamb, M. R., Robertson, L. C. and Knight, R. T. Component mechanism underlying the processing of hierarchically organized patterns: interferences from patients with unilateral cortical lesions. Journal of
Experimental Psychology: Learning, Memory and Cognition 16, 471-483, 1990. 15. Markus, H. S. and Boland, M. "Cognitive activity" monitored by non-invasive measurement of cerebral blood flow velocity and its application to the investigation of cerebral dominance. Cortex 28, 575-581, 1992. 16. Martin, M. Hemispheric specialization for local and global processing. Neuropsychologia 17, 33-40, 1979. 17. Martin, M. Local and global processing: the role of sparsity, Memory and Cognition 7, 476-484, 1979. 18. Navon, D. Forest before trees: the precedence of global features in visual perception. Cognitive P.s3'chology 9, 353 383, 1977. 19. Ribs, F., Gutbrod, K., Gutbrod, B., Steiger, H.-J., Sturzenegger, M. and Mattle, H. P. Determination of cognitive hemispheric dominance by "stereo" transcranial Doppler sonography. Stroke 26, 70-73, 1995. 20. Robertson, L. C. and Delis, D. C. "Part-whole" processing in unilateral brain damaged patients: dysfunction of hierarchical organization. Neuropsychologia 24, 363-370, 1986. 21. Robertson, L. C. and Lamb, M. R. Neurospsychological contributions to part-whole organization. Cognitive Psychology 23, 299-332, 1991. 22. Robertson, L. C., Lamb, M. R. and Knight, R. T. Normal global local analysis in patients with dorsolateral frontal lobe lesions. Neuropsychologia 29, 959-967, 1991. 23. Robertson, L. C., Lamb, M. R. and Zaidel, E. Interhemispheric relations in processing hierarchical patterns: evidence from normal and commissurotomized subjects. Neuropsychology 7, 325-342, 1993. 24. Schnittger, C., Johannes, S. and Mante, T. F. Transcranial Doppler assessment of cerebral blood flow velocity during visual spatial selective attention in humans. Neuroscience Letters 214, 41-44, 1996. 25. Sergent, J. The cerebral balance of power: confrontation or cooperation. Journal of Experimental
Psychology, Human Perception and PerJormance 8, 253-272, 1982. 26. Weekes, N. Y., Carusi, D. and Zaidel, E. Interhemispheric relations in hierarchical perception: a second look. Neuropsychologia 35, 37-44, 1997.
~ )
Pergamon
PII: S0028 3932(97)00038 9
35. No. 8, pp. 1181 1184, 1997 ~', 1997 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0028 3932 9 7 $17.00 + 0.00
Note Blood flow velocity changes in the middle cerebral artery induced by processing of hierarchical visual stimuli CORNELIA SCHNITTGER,* SONKE JOHANNES,* ANOUSCHEH ARNAVAZ* and THOMAS F. MUNTEt *Department of Neurology, Medizinische Hochschule, Hannover, Germany; tDepartment of Cognitive Science, University of California at San Diego, La Jolla, CA 92093-0515, U.S.A. (Received 31 July 1996; accepted 22 February 1997)
Abstraet--Transcranial Doppler ultrasound was used to measure mean flow velocities in both middle cerebral arteries while 16 young subjects performed a visual task involving the processing of hierarchically structured stimuli. Specifically, large (global) letters composed of smaller (local) letters were presented, with the subjects' task being to attend either to the local or to the global level and press a button whenever a target on the designated level occurred. Each run was comprised of a 35-sec period of passive stimulation, followed by 65 sec of active task. A highly significant increase of blood flow was detected upon initiation of the active task, which was clearly present after ca. 4 sec, The flow velocity reached a maximum after 20 sec and remained stable for the remainder of the active condition. No hemispheric differences with respect to global or local conditions were observed, i : 1997 Elsevier Science Ltd Key Words: blood flow velocity; global/local; transcranial Doppler; hemisphere.
for global/local processing, while the evidence for a role of interhemispheric integration for global interference was contradictory [23, 26]. On the other hand, lesion studies in human patients suggested that the right hemisphere subserves the processing of global and the left hemisphere subserves the processing of local information [13, 14, 20, 22]. The temporoparietal junction area has been implicated as being essential for the processing of global and local level information. In several event-related brain potential studies our group has found evidence for parallel processing of global and local level information ill partially lateralized brain systems [6], with a more recent study revealing larger effects of" target detection at left hemispheric recording sites [8]. The present study sought to further investigate the possible lateralization of global/local processing by assessing changes in blood flow of the middle cerebral arteries with transcranial Doppler ultrasound (TCD), while subjects selectively attended to global and local levels of hierarchical visual stimuli. TCD is a non-invasive method with an excellent time resolution with respect to other techniques for the assessment of cerebral blood flow, such as positron emission tomography and SPECT, permitting the monitoring of blood flow velocity changes over a few seconds bilaterally. Assuming that the diameter of the large basal cerebral arteries remains stable, any change in velocity will reflect a change in flow [7, 10]. Several TCD studies have already suggested that diameter changes in the small resist-
Introduction Visual stimuli can be conceptualized as being comprised of information on different levels (e.g., a face is made up from features like eyes, nose, mouth, etc.), which have been termed global and local levels. These levels are thus defined by their relative place in a hierarchy [9, 11, 12, 16-18]. There have been a number of studies suggesting a hemispheric specialization of the processing of global and local level information. For example, when hierarchically constructed stimuli were presented to the visual half-fields of normal subjects a left visual field reaction time advantage was found for the identification of larger letters. while smaller letters were processed faster in the right visual field [25]. Other authors failed to find this left visual half-field advantage in reaction time for identification of larger letters in normal subjects [16] or did not find any visual field asymmetries corresponding to the identification of global or local letters [1, 2]. Studies with commisurotomized subjects using the same technique did not find a specialization of the two hemispheres
t Address for correspondence: Department of Cognitive Science, University of California at San Diego, La Jolla, CA 920930515, U.S.A.; e-mail: tmuente(a~cogsci.ucsd.edu; tel.: + 1-619534-2440; fax: + 1-619-534-1128. 1181
1182
Note
ance vessels caused by metabolic changes due to cerebral activation resulted in velocity changes in the corresponding larger arteries (e.g., [3-5, 15, 19, 23]).
Methods
Subjects Sixteen right-handed healthy volunteers (mean age 25 years, range 20-33 years, eight women) participated in the experiment. None had a family history of left-handedness. All subjects had normal or corrected-to-normal vision.
errors. During the first 35 sec of a run they passively viewed the stimuli. At this time the letter 'H' or 'S' appeared in the middle of the screen and indicated which target letter had to be responded to for the remaining 65 sec of the run (e.g., 'global S' or 'local H'). Forty runs of each attention level were administered during two experimental sessions. Subjects had to respond to target stimuli by pressing a button held in the right or left hand (hands changed every 10 runs). They sat with their head resting against the back of a reclining chair. Pulse rate was measured continuously and blood pressure was obtained at the beginning and the end of each block, consisting of 10 runs, to ensure that significant variations of these parameters did not occur. Response speed and accuracy were assessed as well. All behavioral and TCD measures were analysed with repeated measures analysis of variance (ANOVA).
Equipment Results A commercially available Doppler unit (Multi Dop X 4, DWL Elektronische Systeme GmbH, Sipplingen, Germany) was employed allowing on-line as well as off-line calculation of mean flow velocity (Vmo,n). During the recording session, markers were set automatically at the beginning and the end of the stimulation period to identify relevant episodes of cognitive stimulation. The two Doppler probes were attached to the right and left temporal ultrasound windows above the zygomatic arch by using a special probe holder. Both middle cerebral arteries were insonated at a depth of 45-55 mm depending on the best signal-to-noise ratio.
Reaction times to targets on the global and local levels were 407msec (S.D. 26msec) and 392msec (S.D. 32msec) ( P < 0.003), respectively. The hit rate was 93.1% (S.D. 6.1%) for targets on the global level and 92.0% (S.D. 6.2%) on the local level (n.s.). No significant changes of heart rate or blood pressure were observed during the task. Figure 2A and B illustrates the attention-related changes in (A)
115
Stimuli and procedure The stimuli consisted of the white letters 'H', 'E', 'S' and 'A' presented in the middle of a black video screen. Global letters were constructed from the appropriate placement of local letters, for example a large H made from small Es in a 5 x 5 matrix (see Fig. 1). Global letters were 7.4 times as tall as local letters. The height-to-width ratio of both global and local letters was 3:2. At the viewing distance of 70cm, the global letters subtended 3.7 c'of visual angle in height. The stimuli were flashed in a rapid sequence with a stimulus duration of 100 msec, and an interstimulus interval varying randomly between 800 and 1200 msec. Before each run, subjects were informed on the level (global/local) to be attended during the run. They were also instructed to respond as quickly as possible while minimizing
110
_o o
105
100
95 0
20
40
60
80
(B)
[
115
I ~
Local right MCA [
/
I
110t
E E E E EEEEE E E E E Fig. 1. Example of the stimuli used in the study. A large letter (global level) is made up from smaller letters (local level). The subjects had to attend either to the global or the local level.
100
Time (see)
[
I 0
20
40
60
80
100
Time (sec) Fig. 2. Blood flow velocity in the left and right middle cerebral arteries (passive stimulation baseline = 100%). (A) Mean blood flow response (16 subjects) for the runs with attention to the global level. (B) The same for attention to the local level. In both cases a steep increase for the first few seconds is seen, followed by a plateau phase. Termination of the task causes a brief overshoot of flow, followed by a return to baseline levels.
Note mean flow velocities. A marked and immediate increase of blood flow velocity was observed at the beginning of the active condition. Compared to the level of flow velocity during passive stimulation, the flow increase was 7.1-9.8% for the right and 6.5 8.4% for the left middle cerebral artery for the individual subjects. The change in flow velocity became apparent within the first 4sec of active viewing, reached a steady state after approximately 20 sec and remained relatively constant for the duration of the task. There is a small overshoot preceding the decrease at the end of active stimulation, which is present in most silently performed tasks, and the maximum decrease of flow velocity is reached approximately 16 sec later, followed by an increase to passive stimulation steady state. The statistics were performed in several steps.
E2l~,cts during the.first 20 sec First, an ANOVA including the 10 2-sec periods of the first 20 sec of the active tasks as levels of a factor showed a highly significant increase of blood flow over these periods [F(9,135)= 12.44, P<0.0001], while no main effects were seen for the factors level [global vs local, F(1,15)=1.17, n.s.] and artery [left vs right, F(1,15)=1.87, n.s.]. Neither the levelx artery nor the level x artery x period interactions approached significance. To assess the onset time of the blood flow response in this task, pairwise comparison of the 2-sec intervals with a 2-sec period during the passive stimulation baseline were done, revealing significant effects for all 2-sec periods except for the 0-2 sec interval [all F(I,15)>7.1, all P<0.03]. As the physical stimulation was identical during the passive stimulation baseline and the active stimulation phase, the blood flow response attributable purely to the task occurs within the first 4 sec.
EIl~'cts durhtg the entire recording epoch An ANOVA using the 10 successive 10-sec time intervals of the entire experimental run (see Fig. 2) as levels of a factor revealed a highly significant effect for this factor [F(9,135) = 31.34, P < 0.0001 ]. The factor level approached significance, with the overall flow velocity in the local condition being slightly faster than in the global condition [F(I, 15)= 2.44, P = 0.13]. Neither the factor artery nor any of the interaction effects reached significance. The pairwise comparisons of the single 10-sec periods with the 50-60-sec period (i.e. right before the end of the active task) revealed no significant differences for the 10-20, 20 30, 30-40 and 40 50-sec periods (all F<2.5, n.s.). However, there were significant differences between the 50 60sec and the 0-10-sec intervals [F(1,15)=9.86, P<0.01] and the 60 70-sec interval [F(1,15)=7.18, P<0.03]. Similarly, there was a difference between the 50-60-sec interval and the last two intervals (80 90 and 90-100 sec) during the passive rest [both F(1,15)> 16. P<0.005].
Discussion The current study had two major goals: (i) to demonstrate the feasibility of the TCD method for the assessment of the time course of changes in blood flow related to visual attention, and (ii) to show differential blood flow changes in the two middle cerebral arteries as a function of attention to either global or local features of a visual stimulus. Regarding the first goal, clear evidence of a rapid increase in cerebral blood flow was provided when subjects changed from passive to active viewing of the visual stimuli. In fact, a sig-
1183 nificant increase of blood flow velocity was found within the first 4sec of the active condition. The blood flow reached a stable plateau at about 20 sec for the rest of the run. At the termination of the active viewing period, there was an overshoot of flow velocity with respect to this plateau. This is in line with previous investigations using TCD in cognitive tasks that showed a similar overshoot of blood flow velocity alter termination of the task (e.g., [24]). With respect to the second point of interest, no evidence was found for a differential lateralized change in blood flow as a function of attending to either global or local levels of a stimulus. Several considerations might help to understand this result. The first regards the sensitivity of the Doppler technique. It might be argued that any asymmetries in blood flow change were too small to be detected with TCD. The middle cerebral artery is the main vessel supplying the brain and it might well be that a circumscribed asymmetric blood flow change in a branch of this artery went by undetected in TCD. It is of interest in this respect to review the data of Hartje et al. [5], who assessed blood flow in the middle cerebral artery with TCD during three verbal and three non-verbal tasks. They were able to show a significant asymmetry towards the left middle cerebral artery in all three verbal tasks, while only one non-verbal task (picture matching) led to a significant asymmetry towards the right. The two other non-verbal tasks (cube comparison, figure assembly) that were thought to tax the right hemisphere more than the left on the basis of previous reports did not show a significant difference. Thus, from the data of Hartje et al. [5], it can be concluded that the TCD technique is, in principle, capable of detection of asymmetries in blood flow of the middle cerebral artery. However, in some instances in which asymmetries are expected, these are not observed with TCD. This might, however, be partly due to the fact that both hemispheres contribute to a particular task. This latter line of reasoning might also apply to the current pattern of data. In fact, there is evidence that the processing of such stimuli depends, to a large degree, on the conjoined activity of both hemispheres. For example, Robertson and Lamb [21] have presented a model that envisions the coordinated efforts of four subsystems in the processing of global and local information: (i) a global processor located in the right posterior superior temporoparietal cortex, (ii) a local processor associated with the homologous left hemisphere region, (iii) a bilateral parietal lobe system controlling the distribution of the attentional resources to the global and local levels, and (iv) posterior temporal temporal pathways used to communicate between the subsystems and to fine tune them. One can inlerpret the recent data of Weekes et al. [26] in a similar vein; they found no evidence for a specialization of the hemispheres of global/local information in two commissurotomized subjects. It might well be that the data obtained from patients with lesions in the temporoparietal region, which suggested a differential capability of the two hemispheres for the processing of either global or local information, might reflect the selective breakdown of a bihemispheric system that, when intact, utilizes both hemispheres to a roughly equal extent. Thus, the current TCD data might adequately reflect the state of hemispheric involvement in the global local task after all, namely a largely bilaterally distributed multicomponent processing system. To summarize, the TCD technique appears to be a promising tool in the investigation of cerebral blood flow during cognition. Probably more important than aspects of lateralization, the TCD technique appears capable of capturing the temporal dynamics of blood flow during a cognitive task. An important step forward would be the demonstration of behavioral changes as a function of blood flow changes. Experiments to that effect are currently under way in our laboratory. Acknowled#ements- We thank J. Kilian tbr excellent technical
1184 support, and Berdieke Wieringa, Karin Weissenborn and Mike Matzke for helpful discussions. T.F.M. was supported by a grant from the Hermann and Lilly Schilling Foundation (TS 013/177/96).
References
1. Abivasatos, B. and Wilding, J. Hemispheric differences in matching Stroop-type letter stimuli. Cortex 18, 5-22, 1982. 2. Boles, D. B. Global versus local processing: is there a hemispheric dichotomy? Neuropsychologia 22, 445 455, 1984. 3. Droste, D. W., Harders, A. G. and Rastogi, E. Two transcranial Doppler studies on blood flow velocity in both middle cerebral arteries during rest and the performance of cognitive tasks. Neuropsychologia 27, 1221-1230, 1989. 4. Harders, A. G., Laborde, G., Droste, D. W. and Rastogi, E. Brain activity and blood flow velocity changes: a transcranial Doppler study. International Journal of Neuroscience 47, 91-102, 1988. 5. Hartje, W., Ringelstein, E. B., Kistinger, B., Fabianek, D. and Willmes, K. Transcranial Doppler ultrasonic assessment of middle cerebral artery blood flow velocity changes during verbal and visuospatial cognitive tasks. Neuropsychologia 32, 1443-1452, 1994, 6. Heinze, H. J. and Mtinte, T. F. Electrophysiological correlates of hierachical stimulus processing: dissociation between onset and later stages of global and local target processing. Neuropsychologia 31, 841 852, 1993. 7. Huber, P. and Handa, J. Effects of contrast material, hypercapnia, hyperventilation, hypertonic glucose and papaverine on the diameter of the cerebral arteries. Investigative Radiology 2, 17 32, 1967. 8. Johannes, S., Wieringa, B. M., Matzke, M. and Mante, T. F. Hierarchical visual stimuli: electrophysiological evidence for separate left hemispheric global and local processing mechanisms. Neuroscience Letters 210, l 11-114, 1996. 9. Kinchla, R. A. and Wolfe, J. M. The order of visual processing: "Top-down", "Bottom-up", or "Middleout". Perception and Psychophysics 25, 225 231, 1979. 10. Kontos, H. A. Validity of cerebral arterial blood flow calculations from velocity measurements. Stroke 20, l 3, 1989. 11. Lamb, M. R. and Robertson, L. C. The processing of hierarchical stimuli: effects of retinal locus, locational uncertainty, and stimulus identity. Perception and Psychophysics 44, 172 181, 1988. 12. Lamb, M. R. and Robertson, L. C. Do response
Note time advantage and interference reflect the order of processing of global- and local-level information? Perception and Psychophysics 46, 254-258, 1989. 13. Lamb, M. R., Robertson, L. C. and Knight, R. T. Attention and interference in the processing of global and local information: effects of unilateral temporal parietal junction lesions. Neuropsychologia 27, 47 l 483, 1989. 14. Lamb, M. R., Robertson, L. C. and Knight, R. T. Component mechanism underlying the processing of hierarchically organized patterns: interferences from patients with unilateral cortical lesions. Journal of
Experimental Psychology: Learning, Memory and Cognition 16, 471-483, 1990. 15. Markus, H. S. and Boland, M. "Cognitive activity" monitored by non-invasive measurement of cerebral blood flow velocity and its application to the investigation of cerebral dominance. Cortex 28, 575-581, 1992. 16. Martin, M. Hemispheric specialization for local and global processing. Neuropsychologia 17, 33-40, 1979. 17. Martin, M. Local and global processing: the role of sparsity, Memory and Cognition 7, 476-484, 1979. 18. Navon, D. Forest before trees: the precedence of global features in visual perception. Cognitive P.s3'chology 9, 353 383, 1977. 19. Ribs, F., Gutbrod, K., Gutbrod, B., Steiger, H.-J., Sturzenegger, M. and Mattle, H. P. Determination of cognitive hemispheric dominance by "stereo" transcranial Doppler sonography. Stroke 26, 70-73, 1995. 20. Robertson, L. C. and Delis, D. C. "Part-whole" processing in unilateral brain damaged patients: dysfunction of hierarchical organization. Neuropsychologia 24, 363-370, 1986. 21. Robertson, L. C. and Lamb, M. R. Neurospsychological contributions to part-whole organization. Cognitive Psychology 23, 299-332, 1991. 22. Robertson, L. C., Lamb, M. R. and Knight, R. T. Normal global local analysis in patients with dorsolateral frontal lobe lesions. Neuropsychologia 29, 959-967, 1991. 23. Robertson, L. C., Lamb, M. R. and Zaidel, E. Interhemispheric relations in processing hierarchical patterns: evidence from normal and commissurotomized subjects. Neuropsychology 7, 325-342, 1993. 24. Schnittger, C., Johannes, S. and Mante, T. F. Transcranial Doppler assessment of cerebral blood flow velocity during visual spatial selective attention in humans. Neuroscience Letters 214, 41-44, 1996. 25. Sergent, J. The cerebral balance of power: confrontation or cooperation. Journal of Experimental
Psychology, Human Perception and PerJormance 8, 253-272, 1982. 26. Weekes, N. Y., Carusi, D. and Zaidel, E. Interhemispheric relations in hierarchical perception: a second look. Neuropsychologia 35, 37-44, 1997.