Annals of Behavioral Medicine

, Volume 41, Issue 2, pp 235–242 | Cite as

Self-regulation of Cerebral Blood Flow by Means of Transcranial Doppler Sonography Biofeedback

  • Stefan Duschek
  • Daniel Schuepbach
  • Anselm Doll
  • Natalie S. Werner
  • Gustavo A. Reyes del Paso
Original Article

Abstract

Background

Transcranial Doppler sonography (TCD) allows the continuous non-invasive assessment of intracranial blood flow velocities with high temporal resolution. It may therefore prove suitable for biofeedback of cerebral perfusion.

Purpose

The study explored whether healthy individuals can successfully be trained in self-regulation of cerebral blood flow using TCD biofeedback.

Methods

Twenty-two subjects received visual feedback of flow velocities in the middle cerebral arteries of both hemispheres. They were randomly assigned to two groups, one of which attempted to increase, the other to decrease the signal within eight training sessions. Heart rate and respiratory frequency were also monitored.

Results

Both groups achieved significant changes in flow velocities in the expected directions. Modulations in heart rate and respiratory frequency during biofeedback did not account for these effects.

Conclusions

TCD biofeedback enables efficient self-regulation of cerebral blood flow. It is promising in applications such as the treatment of migraine and post-stroke rehabilitation.

Keywords

Biofeedback Cerebral blood flow Doppler sonography Heart rate Respiration 

Introduction

A variety of physiological mechanisms contribute to the maintenance of adequate cerebral blood flow and its adjustment to situational requirements. In order to ensure stable perfusion of neural tissue, cerebral arterioles constrict during increases and dilate during reductions in cerebral perfusion pressure. These autoregulatory processes buffer fluctuations in systemic hemodynamics, thereby preventing brain ischemia on the one hand and capillary damage and edema formation on the other [1]. Furthermore, cerebral blood flow is perpetually adapted to changing demands resulting from brain activity. As a result of the augmented metabolic rate of the nerve cells, neural activation leads to dilation of cerebral resistance vessels followed by increased blood flow in the active tissue [2, 3]. In addition to flow metabolism coupling, neural mechanisms are involved in blood flow modulation during cerebral activation. A fast-acting neural system directly triggers vasomotor changes in cortical microvessels as a response to activation and deactivation of the brain stem [4, 5].

Research on self-regulation of cerebral blood flow by means of biofeedback still remains sparse. Studies from the field of neurofeedback showed that brain activity may be intentionally modulated by providing subjects with feedback from blood oxygenation level using functional near-infrared imaging [6, 7]. This method, referred to as “hemoencephalography”, was proposed as a therapeutic measure, for instance, in attention-deficit hyperactivity disorder [8], migraine [9], autistic spectrum disorder [10], and deficits in executive skills [11]. Functional magnetic resonance imaging (fMRI) was applied, providing individuals with feedback from the local blood oxygen level-dependent (BOLD) response in specific brain areas. A main problem of this technique is related to the rather low time resolution of fMRI. Due to the inertness of the BOLD signal and limitations in data processing, delays of approximately 60 s occurred between acquisition of the magnetic resonance image and feedback presentation in the first pilot studies [12, 13]. Since self-regulation is most efficient when the delay is minimal [14], brain computer interfaces were developed aiming at immediate feedback of the BOLD signal. Weiskopf et al. [15] presented an interface based on “real-time” fMRI in which the BOLD response could be fed back with a delay of less than 2 s from image acquisition. Using this system, subjects were trained to increase their local BOLD signal in the right anterior insular cortex [16] and right inferior frontal gyrus [17], resulting in behavioral changes in terms of enhanced emotional processing and language-related performance. Despite this progress, the technical arrangement of fMRI is complex and expensive, and the procedure is stressful for most individuals. Therefore, practical applications of fMRI-based biofeedback, for instance in clinical settings, are subject to strict limitations.

Transcranial Doppler sonography (TCD) enables the continuous non-invasive assessment of flow velocities in the basal cerebral arteries, i.e., the anterior, middle, and posterior cerebral arteries [18]. The physical basis of the technique is a frequency shift (“Doppler effect”), which results from relative movement between the source and the receiver of oscillations of almost any type. In Doppler sonography, it is the frequency shift that occurs when ultrasound signals are reflected by the erythrocytes in the blood stream that is measured. The size of the shift is directly proportional to blood flow velocity. Modern engineering makes it possible to simultaneously record flow velocities in corresponding arteries of the left and right hemispheres. Due to its excellent time resolution, the method is superior to fMRI in the assessment of the dynamic component of blood flow required by biofeedback [19, 20]. Unlike the diameters of the cerebral resistance vessels, those of the basal arteries, which are insonated by TCD, remain virtually unchanged under varying conditions of brain activity [21, 22]. Therefore modulations of flow velocities in these arteries do not result from their own vasomotor activity, but reflect blood flow changes in their perfusion territories. A large number of studies have demonstrated that TCD constitutes an excellent tool for quantification of cerebral blood flow modulations related to various psychological processes [19]. In addition to high temporal resolution, advantages arise from the relatively simple technical arrangement and low costs, which could make TCD a suitable tool for biofeedback applications aiming at self-regulation of cerebral blood flow.

The treatment of migraine headaches may be a promising clinical application of TCD-based biofeedback. A number of studies in migraine patients provided evidence for increased cerebral blood flow during the interictal interval. Using TCD, increased flow velocities have been documented in the basal cerebral arteries under resting conditions [23, 24, 25, 26]. In addition, interictal migraineurs showed increased cerebral blood flow reactivity, for instance exaggerated phasic rises in posterior cerebral artery flow velocities during visual stimulation [27, 28]. Enhanced cerebral perfusion results partly from cortical hyperactivity, which has repeatedly been described in migraine (c.f. 29). Furthermore, its genesis relates to vascular dysregulation particularly increased vasomotor reactivity and impaired cerebral autoregulation [30, 31]. As to the pathogenesis of migraine, it is assumed that both cortical sensitization and aberrant cerebrovascular control during the interictal interval are involved in triggering the occurrence of headaches [32, 33]. A biofeedback procedure reducing interictal cerebral blood flow may therefore prove helpful in decreasing the probability of a migraine attack.

The present study explored whether healthy volunteers can successfully be trained to self-regulate cerebral blood flow velocities by means of TCD-based biofeedback. For this purpose, an experimental design was applied, which included attempts both to increase and to decrease flow velocities in the middle cerebral arteries. Ultrasound signals simultaneously obtained from both arteries were visually fed back. The middle cerebral arteries were chosen, because they comprise the largest perfusion territory among the basal cerebral arteries, and a large number of studies have proved that flow modulations in these vessels are sensitive to psychological processes (c.f. 19). In order to identify potential influences of peripheral physiological mechanisms on changes in flow velocities, heart rate and respiratory frequency were monitored. Even though systemic hemodynamics and cerebral blood flow are controlled by different physiological systems, recent research has challenged their complete independence [34, 35, 36]. Respiration affects cerebral perfusion via shifts in CO2 tension and periarteriolar pH, i.e., hypercapnia causes vasodilation with blood flow increase, and hypocapnia causes vasoconstriction [1].

Methods

Participants

Twenty-two university students participated. They were randomly assigned to one of two study groups who were instructed either to increase or decrease blood flow in the middle cerebral arteries. Both groups included seven women and four men. Information regarding age, body mass index (kg/m2), as well as blood pressure, heart rate, respiratory frequency and resting flow velocities in both middle cerebral arteries were gathered in the first training session (see “Procedure”) and are presented in Table 1. None of the participants suffered from a relevant physical disease or mental disorder. Health status was assessed by means of a questionnaire covering diseases of the cardiovascular, respiratory, gastro-intestinal and uro-genital systems, the thyroid, the liver, as well as metabolic diseases and psychiatric disorders. None of the participants used any kind of medication affecting the cardiovascular or central/peripheral nervous system.
Table 1

Age, Body Mass Index, blood pressure, heart rate, respiratory frequency and middle cerebral artery resting blood flow velocities in both experimental groups (all group differences n.s.)

 

Blood flow increase

Blood flow decrease

M

SD

Min

Max

M

SD

Min

Max

Age (years)

22.7

2.4

22

28

24.6

3.4

21

31

Body Mass Index (kg/m2)

20.9

1.6

18.7

23.5

21.6

2.1

18.7

24.5

Systolic blood pressure (mmHg)

109.7

13.8

89

129

115.4

9.7

99

128

Diastolic blood pressure (mmHg)

75.2

9.8

58

90

72.8

6.1

61

83

Heart rate (beats/min)

76.6

10.3

60

96

69.7

9.6

51

81

Respiratory frequency (cycles/min)

14.7

3.0

9.5

21.0

16.4

2.0

13.5

19.0

Resting flow velocity left MCA (cm/s)

65.5

11.4

44.9

83.8

66.8

10.0

48.9

83.6

Resting flow velocity right MCA (cm/s)

62.4

10.6

52.2

82.9

66.5

6.8

54.7

80.0

M means, SD standard deviations, Min minimal values, Max maximal values, MCA middle cerebral artery

Signal Acquisition

A commercially available device (Multidop L2, DWL Elektronische Systeme, Sipplingen, Germany) was applied for TCD. Both middle cerebral arteries were simultaneously insonated through the temporal bone windows, using two 2-MHz transducer probes (for technical details c.f. 19). After vessel identification, the probes were fixed to the head with a tight rubber band. The ultrasound signal was acquired at a depth ranging from 48 to 50 mm in all subjects. The spectral envelope curves of the Doppler signal were analogously exported and digitized at a sample rate of 250 Hz using an analog digital converter (Model ANL-947, MED Associates Inc., St. Albans, VT, USA).

For further data processing a software program (QBasic) specifically designed for the current study was applied. The program allowed on-line integration of the blood flow data over each cardiac cycle and computation of a “mean flow velocity index”. Among the measures typically used in TCD, this index is the least susceptible to artifacts and demonstrates the highest correlation with the blood volume flowing through a vessel [37, 38]. Mean flow velocity values from the left and right middle cerebral arteries were averaged and fed back to the subjects. Respiration was continuously monitored with a Biopac 150 system (Biopac Systems Inc., Goleta, CA, USA) using a piezoelectric thoracic belt (sampling rate, 500 Hz).

Biofeedback Arrangement

On the computer screen the on-line signal of bilateral blood flow velocities was represented by a blue vertical bar (maximum length, 18 cm). The length of the bar increased and decreased according to relative changes in flow velocities. Left of the bar, a numeric scale with the endpoints 0 and 10 was shown. The midpoint of the scale (value, 5) corresponded to average flow velocity recorded during baseline conditions. Increases or decreases by one point corresponded to relative changes in flow velocity by 3% with respect to baseline.

The training consisted of eight sessions, each of which included six biofeedback trials of 4-min duration. All trials were preceded by 2-min baseline periods. Average flow velocity of each baseline was automatically computed and used as a reference (scale point, 5) in the following biofeedback trial. The scale was adjusted to the immediately preceding baseline in each trial because of the well known slow spontaneous oscillations in cerebral blood flow velocities [39]. A pilot study in which a single baseline was used for five trials showed that in some cases spontaneous oscillations led to reference values rather distant from the midpoint of the scale and thus confused the subjects.

Procedure

Sessions were repeated with mean intervals of 2.52 days (SD = 0.53) in the blood flow increase group and 2.38 days (SD = 0.68) in the blood flow decrease group. At the beginning of the first session subjects gave informed consent, and demographic and health information was obtained. Furthermore, blood pressure and heart rate were measured using an automatic inflation monitor (Omron M9 Premium, Omron, Schaumburg, IL, USA). The values of respiratory frequency and resting middle cerebral artery blood flow velocities shown in Table 1 were recorded during the first baseline period of the biofeedback procedure.

Participants were informed that the bar on the computer screen represented blood flow velocities in a pair of large cerebral arteries, and that the task consisted of trying to increase/reduce the velocities, i.e. enlarging/shortening the bar and keeping it as long/short as possible. Subjects were not instructed to use any particular strategy to control the blood flow signal, but were asked about the strategies they applied at the end of each session. Respiration was not mentioned in the instruction. During baselines, participants were only asked not to speak and to keep their eyes open. The instruction was first presented in written form and later on discussed in an unstructured manner.

For participation in the eight training sessions subjects received financial remuneration of €65 independent of their success on the biofeedback task.

Data Analysis

The mean flow velocity index was computed from the envelope curves of the ultrasound signal using our own software. Heart rate was derived from the intervals between the peaks of the curve (inter-systolic intervals). The software AcqKnowledge 3.8.1 was applied for peak detection in the respiratory signal and computation of respiratory frequency. All parameters were averaged across each of the baseline and biofeedback periods, and relative (percent) changes between baseline and biofeedback periods were determined. These change scores were averaged across the six trials of each session, resulting in one mean score per session and parameter.

Statistical analysis comprised analysis of variance (ANOVA) procedures with experimental group (blood flow increase vs. blood flow decrease) as a between subjects factor and training session (sessions 1–8) as a within subject factor. Dependent variables were the change scores for mean flow velocity, heart rate and respiratory frequency. Post hoc t tests were computed where appropriate. In addition to session-wise comparisons between the experimental groups, single-sample t tests were computed within both groups in order to check which of the change scores significantly differed from 0. Further analysis was carried out to evaluate linear and non-linear intra-subject trends in both groups by means of one-way ANOVAs.

Possible effects of heart rate and respiration on cerebral blood flow were determined using regression analysis with the change scores for heart rate and respiratory frequency as predictors and the change scores for mean flow velocity as a dependent variable. Since different physiological mechanisms may be involved in voluntarily increasing and decreasing cerebral blood flow, separate analyses were conducted for both groups as well as for the entire sample. In order to keep alpha-inflation to a minimum, regression analyses were not computed separately for each session, but the models based on overall change scores aggregated across the eight sessions.

Results

Experimental groups did not differ in terms of age, systolic and diastolic blood pressure, heart rate, respiratory frequency, resting blood flow velocities in the left and right middle cerebral arteries, or mean intervals between training sessions (all t < 1.45; all p > .05).

Figure 1 displays the average change scores for mean flow velocity in the eight training sessions. The participants who attempted to increase blood flow velocity showed exclusively positive values ranging between 2% and 3.5% over all sessions. The group trying to decrease flow velocity exhibited a slight increase in session 1, followed by gradually decreasing values in subsequent sessions down to less than 6% in session 8. The ANOVA on flow velocity revealed a significant main effect of experimental group (F[1, 20] = 13.32; p < .01; partial Eta squared = .40), as well as a significant interaction between group and sessions (F[7, 140] = 2.87; p < .01; partial Eta squared = .26). Post hoc testing yielded significant group differences for sessions 2–8 (all t[20] > 2.43; all p < .05). Single-sample t tests showed that in the increase group, change scores for sessions 2–8 were significantly higher than 0 (all t > 2.81; all p < .05). In the decrease group they were significantly below 0 in sessions 7 and 8 (all t[10] > 2.42; all p < .05). A significant linear trend was found in the decrease group (F[1, 10] = 5.45; p < .05; partial Eta squared = .35).
Fig. 1

Relative changes in middle cerebral artery (MCA) blood flow velocity with respect to baseline in the eight training sessions (bars represent standard errors of the mean)

In Fig. 2 the change scores are presented for heart rate. While positive values between 5% and 8.5% were found in the increase group, the decrease group showed relatively small positive and negative scores. The ANOVA revealed a significant main effect of group (F[1, 20] = 29.64; p < .01; partial Eta squared = .60). Group differences were significant in all sessions (all t[20] > 2.18; all p < .05). In the increase group, all values were significantly above 0 (all t[10] > 3.21; all p < .01), and in the decrease group none of them differed significantly from 0 (all t[20] < 2.20; all p > .05). The trend analysis for heart rate did not reveal significant results.
Fig. 2

Relative changes in heart rate with respect to baseline in the eight training sessions (bars represent standard errors of the mean)

The changes scores for respiratory frequency were overall somewhat higher in the increase than in the decrease group (Fig. 3). The ANOVA however did not yield any significant main effect or interaction, nor were there any significant trends.
Fig. 3

Relative changes in respiratory frequency with respect to baseline in the eight training sessions (bars represent standard errors of the mean)

Regression analyses did not reveal a significant prediction of the flow velocity changes in the middle cerebral arteries by changes in heart rate or respiratory frequency. This holds true for the model computed for the entire sample (R = .63; heart rate—Beta = .43, p > .05; respiratory frequency—Beta = .27, p > .05), as well as those for the increase (R = .52; heart rate—Beta = .67, p > .05; respiratory frequency—Beta = −.43, p > .05) and the decrease groups (R = .58; heart rate—Beta = −10, p > .05; respiratory frequency—Beta = .56, p > .05).

The interviews conducted after each session revealed the following classes of strategies used during biofeedback: (1) mental images of tense and relaxed situations (e.g., “I thought of an oral examination.”, “I thought of myself lying in a hammock.”), (2) imagery of bodily changes (e.g., “I thought of more and more blood streaming through my head.”), (3) visualization of the changes in the biofeedback signal (e.g., “I stretched/compressed the bar in my mind.”), (4) verbal instructions (e.g., “I told myself to completely relax.”), and (5) physical strain and relaxation (e.g., “I was tensing/relaxing my muscles.”). A number of subjects however were not able to name any specific strategies (e.g., “I really do not know what I did, but it seemed to work.”). Respiration was hardly mentioned by the participants.

Discussion

As a main result the present study demonstrated that TCD-based biofeedback enables individuals to voluntarily increase and decrease blood flow velocities in the basal cerebral arteries. Given that the diameters of the basal arteries virtually remain constant during varying brain activity [21, 22], the blood flow modulations in the middle cerebral arteries can be attributed to constriction and dilation of the resistance vessels in their perfusion territory. The middle cerebral arteries supply the lateral and superior parts of the frontal and parietal cortices, as well as the temporal lobes, the insular cortex and subcortical structures such as the striatum and pallidum [40]. Blood flow changes in these areas are most likely to be related to metabolic processes as well as to direct neurogenic vasomotor control. Mediated by biochemical factors such as K+, H+, and adenosine, arteriolar tone rises and falls with changes in brain metabolism [2]. In neurally mediated blood flow regulation, innervation of the cortical resistance vessels by fibers originating from lower brain areas is well known [4, 41]. These fibers form part of an intracranial neural vaso-modulatory system consisting of cholinergic and serotonergic neurons that project from the ascending reticular activating system (ARAS) to the cortex [5, 42]. In animals, for instance, the stimulation of the nucleus basalis of Meynert of the brain stem was shown to be followed by cortical blood flow increase [43]. Activity of the ARAS is linked to the control of attentional arousal, which is accessible to self-regulation [44]. One may therefore hypothesize that in addition to metabolic processes in the perfusion territory of the middle cerebral arteries, the observed changes in flow velocity related to modulations in brain stem activity and arousal level.

Respiratory frequency tended to rise during the biofeedback procedure, especially in the group that was trained to increase the signal. Respiration has a strong impact on CO2 tension and periarteriolar pH, which in turn modulate arteriolar diameters [1]. Regression analysis however indicated that changes in respiratory frequency did not significantly affect blood flow velocities in any of the study groups. Also, the interviews conducted after the training session suggested that the participants did not use hypo- or hyperventilation as strategies to influence the blood flow signal. Regarding heart rate, no significant change was found in the participants trying to reduce flow velocity, whereas those trained in blood flow enhancement showed a marked increase. One may hypothesize that the latter observation reflects elevated sympathetic tone accompanying the attempt to raise the signal. This would be in accordance with voluntary increase of physical arousal, which some subjects reported using as a strategy to enlarge the bar (e.g., “I tensed up my whole body.”). However, as indicated by the lack of effect in the regression analyses, heart rate acceleration was not transferred to cerebral perfusion. This is consistent with previous observations that only relatively rapid fluctuations in peripheral hemodynamics affect cerebral blood flow, and slower changes are buffered by cerebral autoregulation [34, 35, 36]. Taken together, the data did not reveal evidence for the involvement of peripheral physiological mechanisms in the self-regulation of cerebral blood flow velocities, suggesting that neural mechanism dominated in mediating the effects.

The extent and patterns of changes in cerebral perfusion differed between participants from both training groups. Even though the increase group exhibited a significant blood flow elevation even in the first training session, its magnitude was restricted to 2% to 3% on average in the following sessions. The magnitude of this effect is comparable to the increase in middle cerebral artery flow velocities which is observed during the execution of cognitive tasks involving for instance attention, arithmetic processing or mental planning [20, 35, 36, 42]. The decrease group showed a progressive blood flow reduction in the course of sessions. Though significance was not reached before session 7, the magnitude of the effect is underlined by a linear trend accounting for 35% of the variance in the change scores, as well as by an average reduction by 6% in session 8. Self-regulation of brain perfusion is presumably limited by autoregulatory processes that keep fluctuations in cerebral blood flow to a minimum in healthy individuals. It would appear that voluntary blood flow enhancement triggered more pronounced counter-regulation than did its reduction. Cardiac acceleration as observed in the increase group is accompanied by rising cerebral perfusion pressure, which in turn triggers compensatory arteriolar constriction and thus blood flow reduction [35, 36]. Also, the tendency towards increased respiratory frequency may have limited blood flow enhancement. While decreased respiratory frequency leads to lower periarteriolar pH (acidosis) and compensatory increase in capillary diameters and blood flow, hyperventilation induces blood flow reduction [1].

Obviously, our conclusions about the physiological mechanisms underlying the changes in cerebral blood perfusion must to some degree remain hypothetical. In future studies application of electroencephalography (EEG) measures of cortical activity during the biofeedback procedure may give further insight into the role of neural processes. In addition, a more sophisticated analysis of respiratory and cardiovascular functions would prove helpful. The BOLD signal as well as flow velocities in the middle cerebral arteries are sensitive to fluctuations in arterial CO2 [45]. The recording of respiratory frequency is certainly not sufficient to determine fluctuations in CO2 level, thus our methods were suboptimal in evaluating to which degree the observed changes in cerebral blood flow were due to respiratory modulations. In order to better estimate such effects, more comprehensive respiratory assessment including breath to breath capnographic measurement and recording of respiratory volume are essential in future research. To control for possible effects of cardiovascular changes, continuous registration of blood pressure and cardiac output would be beneficial. One should also not overlook the relatively small size of the present sample, which limited the statistical power of the regression analyses, especially those conducted separately in both study groups.

Psychophysiological treatment of migraine headaches is a tempting application of TCD-based biofeedback. While it has been shown that beneficial effects of relaxation techniques and acupuncture in migraine are accompanied by reduction in intracranial flow velocities [46, 47, 48], cerebral blood flow has hardly been used as a direct target dimension in biofeedback so far [9]. Current biofeedback treatment of migraine aims at either general physical relaxation, reduction of slow cortical potentials or constriction of the superficial temporal arteries using photoplethysmography [49, 50]. While EEG biofeedback addresses the neural aspect of migraine pathology, i.e., interictal cortical hyperactivity, vasoconstriction training aims at reducing the vasomotor aberrations. In the latter case, one should not overlook the fact that the temporal arteries constitute extracranial vessels supplying for instance the face and the pericranium, and that they are not directly involved in the pathogenesis. On the other hand, blood flow velocities in the basal cerebral arteries, which are insonated by TCD, vary according to diameter changes of the intracranial arterioles and capillaries, where the actual pathological processes occur [30, 31]. TCD-based biofeedback may counteract the excessive dilation of cerebral resistance vessels, thereby directly intervening in these processes. Considering the fact that the reduction of intracranial flow velocities by biofeedback are most likely due to neural deactivation, the strategy may also be efficient in counteracting cortical hyperactivity. TCD biofeedback may thus exert beneficial effects on both the neural and vascular components of migraine pathology. In this sense it combines the merits of EEG biofeedback and vasoconstriction training and may prove superior to each of these.

The rehabilitation of ischemic stroke patients may be another field of application of TCD-based biofeedback. In stroke patients, reduced cerebral blood flow has repeatedly been reported, especially diminished increases of blood flow velocities in the basal arteries during cognitive and motor activity [51, 52, 53]. Longitudinal studies identified the magnitude of these blood flow increases as a prognostic factor in post-stroke rehabilitation. In patients with Broca’s aphasia, Silvestrini et al. [52] showed that the flow velocity increase in the left middle cerebral artery during language processing as measured at disease onset predicted the degree of functional recovery after 2 months of speech therapy. Moreover, the success of rehabilitation was greater in patients who experienced an enhancement of the blood flow response in the right middle cerebral artery across the therapy interval. Similarly, the recovery from stroke related memory disorders was found to depend on the magnitude of the cerebral blood flow increase during an object recognition task [53]. Given the importance of cerebral blood flow for post-stroke recovery, TCD-based biofeedback aiming at increasing cerebral flow velocities may be a valuable procedure to optimize the rehabilitation process.

To sum up, the study is the first to provide evidence that biofeedback based on Doppler sonography enables self-regulation of cerebral blood flow. Changes in heart rate and respiratory frequency did not play a relevant role in the self-regulatory process, suggesting that self-regulation of cerebral blood flow was predominantly mediated by mechanisms restricted to the central nervous system. As a limitation of the procedure, one should bear in mind the relatively low spatial resolution of TCD, which is determined by the size of the brain area supplied by the artery under study [19]. While simultaneous bilateral assessment could provide training in hemisphere specific blood flow changes, it is not possible to induce flow modulations restricted to smaller brain areas. In contrast to imaging methods, TCD is relatively easy applicable and much more cost efficient, which would make it a valuable tool for biofeedback applications such as the treatment of migraine and post-stroke rehabilitation.

Notes

Acknowledgements

The cooperation between the Universities of Munich and Jaén was supported by the German Academic Exchange Service and the Spanish Ministry of Education and Science. We are grateful to Stella Bollmann and Tanja Mannhart for their help with the data acquisition and analysis.

Conflicts of Interest Statement

The authors have no conflict of interest to disclose.

References

  1. 1.
    Paulson OB. Blood-brain barrier, brain metabolism and cerebral blood flow. Eur Neuropsychopharmacol. 2002; 12: 495–501.PubMedCrossRefGoogle Scholar
  2. 2.
    Huettel SA, Song AS, McCarthy G. Functional magnetic resonance imaging. Sunderland, MA: Sinauer; 2004.Google Scholar
  3. 3.
    Iadecola C. Neurovascular regulation in the normal brain and in Alzheimer’s disease. Nat Rev Neurosci. 2004; 5: 347–360.PubMedCrossRefGoogle Scholar
  4. 4.
    Sándor P. Nervous control of the cerebrovascular system: Doubts and facts. Neurochem Int. 1999; 35: 237–259.PubMedCrossRefGoogle Scholar
  5. 5.
    Sato A, Sato Y, Uchida S. Regulation of regional cerebral blood flow by cholinergic fibres originating in the basal forebrain. Int J Dev Neurosci. 2001; 19: 327–337.PubMedCrossRefGoogle Scholar
  6. 6.
    Freides D, Aberbach L. Exploring hemispheric differences in infrared brain emissions. J Neurother. 2004; 8: 53–61.CrossRefGoogle Scholar
  7. 7.
    Toomim H, Mize W, Kwong PC, Toomim M, Marsh R, Kozlowski GP, Kimball M, Rémond A. Intentional increase of cerebral blood oxygenation using hemoencephalography (HEG): An efficient brain exercise therapy. J Neurother. 2004; 8: 5–21.CrossRefGoogle Scholar
  8. 8.
    Mize W. Hemoencephalography—a new therapy for attention deficit hyperactivity disorder (ADHD): Case report. J Neurother. 2004; 8: 77–97.CrossRefGoogle Scholar
  9. 9.
    Carmen J A. Passive infrared hemoencephalography: Four years and 100 migraines. J Neurother. 2004; 8: 23–51.CrossRefGoogle Scholar
  10. 10.
    Coben R, Padolsky, IP. Infrared imaging and neurofeedback: Initial reliability and validity. J Neurother. 2007; 11: 3–13.CrossRefGoogle Scholar
  11. 11.
    Sherrill R. Effects of hemoencephalography (HEG) training at three prefrontal locations upon EEG ratios at Cz. J Neurother. 2004; 8: 63–76.CrossRefGoogle Scholar
  12. 12.
    Yoo SS, Jolesz FA. Functional MRI for neurofeedback: Feasibility study on a hand motor task. Neuroreport. 2000; 13: 1377–1381.CrossRefGoogle Scholar
  13. 13.
    Posse S, Fitzgerald D, Goo K, et al. Real-time fMRI of temporolimbic regions detects amygdala activation during single-trial self-induced sadness. NeuroImage. 2003; 18: 760–768.PubMedCrossRefGoogle Scholar
  14. 14.
    Rockstroh B, Elbert T, Birbaumer N, Lutzenberger W. Biofeedback-produced hemispheric asymmetry of slow cortical potentials and its behavioral effects. Int J Psychophysiol. 1990; 9: 151–165.PubMedCrossRefGoogle Scholar
  15. 15.
    Weiskopf N, Veit R, Erb M, et al. Physiological self-regulation of regional brain activity using real-time functional magnetic resonance imaging (fMRI): Methodology and exemplary data. Neuroimage. 2003; 19: 577–586.PubMedCrossRefGoogle Scholar
  16. 16.
    Caria A, Veit R, Sitaram R, et al. Regulation of anterior insular cortex activity using real-time fMRI. Neuroimage. 2007; 35: 1238–1246.PubMedCrossRefGoogle Scholar
  17. 17.
    Rota G, Sitaram R, Veit R, et al. Self-regulation of regional cortical activity using real-time fMRI: The right inferior frontal gyrus and linguistic processing. Hum Brain Mapp. 2009; 30: 1605–1614.PubMedCrossRefGoogle Scholar
  18. 18.
    Aaslid R, Markwalder TM, Nornes H. Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries. J Neurosurgery. 1982; 57: 769–774.CrossRefGoogle Scholar
  19. 19.
    Duschek S, Schandry R. Functional transcranial Doppler sonography as a tool in psychophysiological research. Psychophysiology. 2003; 40: 436–454.PubMedCrossRefGoogle Scholar
  20. 20.
    Schuepbach D, Boeker H, Duschek S, Hell D. Rapid cerebral hemodynamic modulation during mental planning and movement execution: Evidence of time-locked relationship with complex behavior. Clin Neurophysiol. 2007; 118: 2254–2262.PubMedCrossRefGoogle Scholar
  21. 21.
    Kontos HA. Validity of cerebral arterial blood flow calculations from velocity measurements. Stroke. 1998; 20: 1–3.CrossRefGoogle Scholar
  22. 22.
    Giller CA, Bowman G, Dyer H, Mootz L, Krippner W. Cerebral arterial diameters during changes in blood pressure and carbon dioxide during craniotomy. Neurosurgery. 1993; 32: 737–741.PubMedCrossRefGoogle Scholar
  23. 23.
    Thie A, Fuhlendorf A, Spitzer K, Kunze K. Transcranial Doppler evaluation of common and classic migraine. Part I. Ultrasonic features during the headache-free period. Headache. 1990; 30: 201–208.PubMedCrossRefGoogle Scholar
  24. 24.
    Abernathy M, Donnelly G, Kay G, et al. Transcranial Doppler sonography in headache-free migraineurs. Headache. 1994; 34: 198–203.PubMedCrossRefGoogle Scholar
  25. 25.
    Kastrup A, Thomas C, Hartmann C, Schabet M. Cerebral blood flow and CO2 reactivity in interictal migraineurs: A transcranial Doppler study. Headache. 1998; 38: 608–613.PubMedCrossRefGoogle Scholar
  26. 26.
    Arjona A, de Torres LA, Serrano-Castro PJ, Guardado-Santervas PL, Olivares J, Rubi-Callejon J. A transcranial doppler study in interictal migraine and tension-type headache. J Clin Ultrasound. 2007; 35: 372–375.PubMedCrossRefGoogle Scholar
  27. 27.
    Bäcker M, Sander D, Hammes MG, et al. Altered cerebrovascular response pattern in interictal migraine during visual stimulation. Cephalalgia. 2001; 21: 611–616.PubMedCrossRefGoogle Scholar
  28. 28.
    Nedeltchev K, Arnold M, Schwerzmann M, et al. Cerebrovascular response to repetitive visual stimulation in interictal migraine with aura. Cephalalgia. 2004; 24: 700–706.PubMedCrossRefGoogle Scholar
  29. 29.
    Silberstein SD. Migraine pathophysiology and its clinical implications. Cephalalgia. 2004; 24 Suppl 2: 2–7.PubMedCrossRefGoogle Scholar
  30. 30.
    Zaletel M, Strucl M, Bajrovic FF, Pogacnik T. Coupling between visual evoked cerebral blood flow velocity responses and visual evoked potentials in migraneurs. Cephalalgia. 2005; 25: 567–574.PubMedCrossRefGoogle Scholar
  31. 31.
    Vernieri F, Tibuzzi F, Pasqualetti P, et al. Increased cerebral vasomotor reactivity in migraine with aura: An autoregulation disorder? A transcranial Doppler and near-infrared spectroscopy study. Cephalalgia. 2008; 28: 689–695.PubMedCrossRefGoogle Scholar
  32. 32.
    Parsons AA, Strijbos PJLM. The neuronal versus vascular hypothesis of migraine and cortical spreading depression. Curr Opin Pharmacol. 2003; 3: 73–77.PubMedCrossRefGoogle Scholar
  33. 33.
    Dodick DW. Examining the essence of migraine—Is it the blood vessel or the brain? A debate. Headache. 2008; 48: 661–667.PubMedCrossRefGoogle Scholar
  34. 34.
    Zhang R, Zuckerman JH, Giller CA, Levine BD. Transfer function analysis of dynamic cerebral autoregulation in humans. Am J Physiol Heart Circ Physiol. 1998; 274: H233–H241.Google Scholar
  35. 35.
    Duschek S, Werner N, Kapan N, Reyes del Paso GA. Patterns of cerebral blood flow and systemic hemodynamics during arithmetic processing. J Psychophysiol. 2008; 22: 81–90.CrossRefGoogle Scholar
  36. 36.
    Duschek S, Heiss H, Schmidt FH, Werner N, Schuepbach D. Interactions between systemic hemodynamics and cerebral blood flow during attentional processing. Psychophysiology. 2010; 47:1159–1166.PubMedGoogle Scholar
  37. 37.
    Newell DW, Aaslid R, Lam A, Mayberg TS, Winn HR. Comparison of flow and velocity during dynamic autoregulation testing in humans. Stroke. 1994; 25: 793–797.PubMedGoogle Scholar
  38. 38.
    Stroobant N, Vingerhoets G. Transcranial Doppler ultrasonography monitoring of cerebral hemodynamics during performance of cognitive tasks: A review. Neuropsychol Rev. 2000; 10: 213–231.PubMedCrossRefGoogle Scholar
  39. 39.
    Diehl R, Diehl B, Sitzer M, Hennerici M. Spontaneous oscillations in cerebral blood flow velocity in normal humans and in patients with carotid artery disease. Neurosci Lett. 1991; 127: 5–8.PubMedCrossRefGoogle Scholar
  40. 40.
    Haines DE. Neuroanatomy. An atlas of structures, sections, and systems. Philadelphia, PA: Lippincott Williams & Wilkins; 2007.Google Scholar
  41. 41.
    Hamel E, Vaucher E, Tong XK, St-Georges M. Neuronal messengers as mediators of microvascular tone in the cerebral cortex. Int Congr Ser. 2002; 1235: 262–276.CrossRefGoogle Scholar
  42. 42.
    Szirmai I, Amrein I, Pálvögyi L, Debreczeni R, Kamondi A. Correlation between blood flow velocity in the middle cerebral artery and EEG during cognitive effort. Cognit Brain Res. 2005; 24: 33–40.CrossRefGoogle Scholar
  43. 43.
    Biesold D, Inanami O, Sato A, Sato Y. Stimulation of the nucleus basalis of Meynert increases cerebral cortical blood flow in rats. Neurosci Lett. 1998; 98: 39–44.CrossRefGoogle Scholar
  44. 44.
    Posner MI, Petersen, SE. The attention system of the human brain. Ann Rev Neurosci. 1990; 13: 25–42.PubMedCrossRefGoogle Scholar
  45. 45.
    Wise WG, Ide K, Poulin MJ, Tracey I. Resting fluctuations in arterial carbon dioxide induce significant low frequency variations in BOLD signal. Neuroimage. 2004; 21: 1652–1664.PubMedCrossRefGoogle Scholar
  46. 46.
    McGrady A, Wauquier A, McNeil A, Gerard G. Effect of biofeedback-assisted relaxation on migraine headache and changes in cerebral blood flow velocity in the middle cerebral artery. Headache. 1994; 34: 424–428.PubMedCrossRefGoogle Scholar
  47. 47.
    Wauquier A, McGrady A, Aloe L, Klausner T, Collins B. Changes in cerebral blood flow velocity associated with biofeedback-assisted relaxation treatment of migraine headaches are specific for the middle cerebral artery. Headache. 1995; 35: 358–362.PubMedCrossRefGoogle Scholar
  48. 48.
    Bäcker M, Hammes M, Sander D, et al. Changes of cerebrovascular response to visual stimulation in migraineurs after repetitive sessions of somatosensory stimulation (acupuncture): A pilot study. Headache. 2004; 44: 95–101.PubMedCrossRefGoogle Scholar
  49. 49.
    Nestoriuc Y, Martin A. Efficacy of biofeedback for migraine: A meta-analysis. Pain. 2007; 128: 111–127.PubMedCrossRefGoogle Scholar
  50. 50.
    Nestoriuc Y, Martin A, Rief W, Andrasik F. Biofeedback treatment for headache disorders: A comprehensive efficacy review. Appl Psychophysiol Biofeedback. 2008; 33: 125–140.PubMedCrossRefGoogle Scholar
  51. 51.
    Silvestrini M, Troisi E. Matteis M, Cupini LM, Caltagirone C. Involvement of the healthy hemisphere in recovery from aphasia and motor deficiency in patients with cortical ischemic infarction: A transcranial Doppler study. Neurology. 1995; 45: 1815–1820.PubMedGoogle Scholar
  52. 52.
    Silvestrini M, Troisi E, Matteis M, Razzano C, Caltagirone C. Correlations of flow velocity changes during mental activity and recovery from aphasia in ischemic stroke. Neurology. 1998; 50: 191–195.PubMedGoogle Scholar
  53. 53.
    Bragoni M, Caltagirone C, Troisi E, Matteis M, Verzieri F, Silvestrini M. Correlation of cerebralhemodynamic changes during mental activity and recovery after stroke. Neurology. 2000; 55: 35–40.PubMedGoogle Scholar

Copyright information

© The Society of Behavioral Medicine 2010

Authors and Affiliations

  • Stefan Duschek
    • 1
    • 4
  • Daniel Schuepbach
    • 2
  • Anselm Doll
    • 1
  • Natalie S. Werner
    • 1
  • Gustavo A. Reyes del Paso
    • 3
  1. 1.Department of PsychologyUniversity of MunichMunichGermany
  2. 2.Psychiatric University Hospital ZürichZürichSwitzerland
  3. 3.University of JaénJaénSpain
  4. 4.Department PsychologieLudwig-Maximilians-Universität MünchenMunichGermany

Personalised recommendations