Volitional Control of Heartbeat and Its Dependence on Pallium

Abstract

It is believed that we cannot change our heart rhythm by will because the heartbeat is mainly controlled by the autonomic nervous system (ANS), which cannot be affected directly by subjective will. An experiment was designed to determine whether the heartbeat and ANS could be controlled by volition, and, if it is true, how they were controlled. It was demonstrated that the ANS could be partly controlled by volition. The volition, which tended to slow down the heartbeat, initiated synchronized activity in the medial prefrontal cortex, inhibited the sympathetic system, and then decreased the heartbeat. On the other hand, another kind of volition, which sped up the heartbeat, initiated desynchronized activity at the precentral, central, parietal, and occipital regions, inhibited the parasympathetic system and excited the sympathetic system, and then increased the heartbeat. Moreover, information flow from posterior cortex to anterior cortex was observed during the experiment. The parietal area played an important role in triggering the sensorimotor cortex and integrating the information, and the information flow from the central and precentral cortex to heart was dominant. All that demonstrated that volition can partly control the heartbeat, but the behavior was different from the motor nervous system.

Keywords

volition heart rate autonomic nervous system electroencephalogram directed transfer function

Introduction

The skeletal muscle can voluntarily move with a command from the brain, but it is believed that the heart rate (HR) and ANS are tightly controlled homeostatic networks capable of little voluntary control. Recently, research showed that cortical and subcortical brain systems were involved in modulating the autonomic efferent outflow to the heart, and suppression of their activity will impact on cardiovascular arousal.^1-3

Emotion, attention, and stress can affect ANS and heartbeat.^4-8 The learning and reward train of biofeedback is an effective treatment to cardiovascular diseases.^9,10 Qigong and yoga can affect cardiovascular function by relaxation or concentrating the volition on heart.^11-13 Electrical stimulation of the insula, anterior cingulate, cortex, and medial temporal lobe have all demonstrated changes in HR.¹⁴ Direct projections from cortex to brain stem and spinal regions were implicated in control of the cardiovascular system.¹⁵ These findings indicate that it may be possible to control the heartbeat. Our study was designed to determine whether the heartbeat and ANS could be controlled by volition, and what will happen in the pallium when we try to do so.

A variety of maneuvers, such as gambling,¹⁶ mental arithmetic,¹⁷ emotion, and attention^4-8, were used to elevate HR. These studies showed that sympathetic and parasympathetic activities are crucial factors that affect HR. Heart rate variability (HRV) can be used to evaluate parasympathetic and sympathetic activities.^18-21 It was suggested that high frequency (HF) and low frequency (LF) were, respectively, related to parasympathetic and sympathetic activities, and LF/HF had been proposed as an index of sympathovagal balance.

A number of approaches have been proposed to estimate the functional connectivity with electroencephalogram (EEG) signals, such as cross-correlation, spectral coherence, and synchronization.^22-24 These techniques were limited to bivariate time series, because they do not make use of the whole covariance structure for multivariate data. In addition, these methods cannot reflect the direction of the information flow within the functional coupling of EEG rhythms, and were poor at analyzing connectivity between brain and heart. The directed transfer function (DTF) method can overcome this limitation by multivariate autoregression (MVAR).^25,26

The MVAR model has been successfully used for estimating the direction of the corticomuscular information flow.^27,28 Original electrocardiogram (ECG) data cannot be used to investigate the information flow between brain and heart because the autonomic nerve regulates heart performance by changing two R peaks (RR) intervals. It has been found that EEG activity was synchronized to the cardiac cycle and the fluctuation of HRV.²⁹ Therefore, the wavelet packet parameters of EEG and the interval between 2 successive R peaks (RR interval) were used to demonstrate the connectivity between brain and heart. Corticocortical functional coupling during volitional control of the heartbeat was also studied.

Materials and Methods

Subjects

Fifty-four healthy male subjects (22-27 years old) voluntarily participated in the study. The subjects were asked to perform the experiment in which HR was controlled below their normal level (HR_down task). Another task was designed to control the heart above their normal level (HR^up task). The subjects were instructed to avoid alcohol, tea, coffee, and strenuous exercise for 12 hours before the experiment, and they were screened carefully with history and physical examinations. The investigation was carried out with the approval of the Xi’an Jiaotong University Ethics Committee, and informed written consent was obtained from each subject after the experimental procedures had been explained.

Experimental Protocol

Experiments were performed between 7 pm and 10 pm. Both HR_down and HR^up tasks were performed in a quiet laboratory in which the temperature was controlled to 23°C ± 1°C.

Before the experiment, the normal ECG of the subject was recorded, and RR intervals were sent to a computer from which two kinds of auditory tones were produced. The tones respectively corresponded to a 10% decrease and increase of the RR intervals, and used to initiate the volitional command “beat” during the HR_down and HR^up task (Figure 1).

In order to confirm that the changes of HR were mainly affected by volition and eliminate the effects of the auditory tone on subjects, a baseline session was designed. During the baseline, subjects remained listening to the same tone but did not concentrate their mind on the heart.

In the HR_down task, subjects were asked to concentrate their mind on the heart and drive their heart to beat with volitional commands “beat” from the brain. In the HR^up task, the protocol was the same as in HR_down, except that the command followed the quicker auditory notes.

Physiological Measurements

EEG was recorded with the Neuroscan 32 channel system (Neuroscan, El Paso, TX). Cap electrodes were positioned according to the International 10-20 System, and electrode impedances were less than 5 kohm per site. The EEG was recorded at 18 scalp locations: left and right frontal (FP1, FP2), mid-frontal (FC3, FC4, and FCz), central (C3, C4, and Cz), parietal (P3, P4, and Pz), occipital (O1, O2, and Oz), and temporal (FT7, FT8, T7, and T8) regions (Figure 2). All electrodes were referenced to linked ear lobe electrodes.

Surface Ag/AgCl electrodes were attached to the chest for recording the ECG (2 electrodes were attached to the right collarbone and left rib, respectively, and the grounding electrode was placed on the lower right rib), and physiological signals were filtered by bandpass from 0.01 to 100 Hz. The signal was sampled at 500 Hz and digitized at 16 bit.

Data Analysis

HR was estimated from the RR intervals. Cardiac autonomic activity was obtained from power spectral analysis of RR intervals named as HRV. Magnitudes of spectral frequencies are considered indices of sympathetic and parasympathetic tone.^30-32 Spectral power of HRV was calculated in the low-frequency (LF: 0.04-0.15 Hz) and high-frequency (HF: 0.15-0.4 Hz) ranges. The fluctuation of HR in LF range is influenced by the baroreceptor system and reflects sympathetic as well as parasympathetic activities.³³ HF is mainly related to parasympathetic efferent activities. LF/HF ratio reflects sympathetic outflow.^18-20 To minimize the effect of total power on the LF and HF components, the relative values in proportion to the total power in the frequency range from 0.04 to 0.4 were calculated and the power was expressed in normalized units (nuLF and nuHF).³⁴

Scalp EEG provides a wealth of information about human brain dynamics.³⁵ Although EEG is limited with respect to inferring neuroanatomical generators of surface potentials, it has higher temporal resolution than techniques such as positron emission tomography or functional magnetic resonance imaging.³⁶ In the study, the wavelet packet decomposition to EEG was applied to identify cortical responses during the volitional task. Daubechies 10 was adopted as the mother wavelet. After 8-octave wavelet packet decomposition, EEG frequency bands of alpha (8-13 Hz) and beta (14-30 Hz) were obtained. The log transformation (ln) of the wavelet packet energy was used.

DTF, based on MVAR models, shows its usefulness in characterizing causal interaction pattern between neurons, and has been used to estimate the direction of the corticocortical and corticomuscular information flow in the previous experiments.^26,27,37 According to standardized rules by Kaminski and Blinowska,²⁵ the EEG data should be preliminarily normalized by subtracting the mean value and dividing by the variance before computing the DTF. An important step for DTF was the computation of MVAR defined as:

\sum_{k = 0}^{p} A (k) X (t - k) = E (t), A (0) = I,

where, $X (t) = {[x_{1} (t), x_{2} (t), \dots, x_{N} (t)]}^{T}$ represent N channel signals at time t, $E (t) = {[e_{1} (t), e_{2} (t), \dots, e_{N} (t)]}^{T}$ is a vector of zero-mean uncorrelated white noise process, $A (1), A (2), \dots, A (p)$ are $N \times N$ matrices of model coefficients, and p is the model order.

To investigate the spectral properties of the examined process, the MVAR model is transformed to its frequency domain:

A (f) X (f) = E (f)

A (f) = \sum_{k = 0}^{p} A (k) e^{- j 2 π f Δ t k}

$Δ t$ is the temporal interval between 2 samples. The transformed MVAR model can then be rewritten as

X (f) = A^{- 1} (f) E (f) = H (f) E (f)

DTF from the jth to ith channel was defined as²⁵:

γ_{i j}^{2} (f) = \frac{{| H_{i j} (f) |}^{2}}{\sum_{m = 1}^{N} {| H_{i m} (f) |}^{2}} .

DTF is a normalized value ranging from 0 to 1. H(f) is a transfer function matrix. Since H(f) is not a symmetric matrix, the difference between $γ_{i j} (f)$ and $γ_{j i} (f)$ was an asymmetric information flow between the jth and ith channels, namely the “direction” of this flow.

The model order was 7, as estimated by the Akaike criterion suggested in previous DTF studies.^25,37 EEG data of 18 electrodes were simultaneously given as input to the MVAR model. Directionality of the functional cortical coupling among all the pair combinations of these electrodes at alpha and beta bands was computed. RR intervals, and wavelet packet parameters of EEG corresponding to each RR interval, were used to compute the directional information flow between brain and heart. The RR intervals, and corresponding wavelet packet parameters, were interpolated at 1-second intervals by cubic spline interpolation, and then as an input to the MVAR model (in this case N = 2) to calculate the DTF values. Results were expressed as DTF difference and denoted by DTFdiff.

Paired t test was performed with Sigma Plot. P < .05 was considered statistically significant and data were presented as mean ± standard error of the mean.

Results

HR and HRV

Compared with the baseline, HR, LF/HF, and nuLF significantly decreased (P < .01), whereas HF, nuHF, and total power significantly increased (P < .01) in volition session for HR_down task (Table 1). The results revealed that volition can make the heartbeat slower by decreasing sympathetic and strengthening parasympathetic activity.

Table 1.

Changes of HR and HRV With Volition in HR_down Task.^a

	HR (beats/min)	TP (ms²)	HF (ms²)	LF (ms²)	nuHF	nuLF	LF/HF
Volition	73.0 ± 2.1**	303.0 ± 55.9*	146.0 ± 32.0**	144.0 ± 24.7	46.9 ± 2.7**	53.1 ± 2.7**	1.4 ± 0.2**
Baseline	75.0 ± 2.2	247.0 ± 44.3	106.0 ± 27.5	129.0 ± 18.0	37.7 ± 3.1	62.3 ± 3.1	2.1 ± 0.3

Abbreviations: HRV, heart rate variability; HR, heart rate; TP, total power; HF, high frequency; LF, low frequency; nuHF, normalized high frequency; nuLF, normalized low frequency.

Data are presented as mean ± standard error of the mean.

P < .05 versus baseline (paired t test).

P < .01 versus baseline (paired t test).

In the HR^up task, HR was significantly higher (74 ± 1.6 beats/min) compared with baseline (71 ± 1.5 beats/min). LF/HF and nuLF significantly increased, whereas HF and nuHF significantly decreased (P < .05; Table 2). Sympathetic activity increased and parasympathetic decreased when the heartbeat was quickened by volition.

Table 2.

Changes of HR and HRV With Volition in HR^up Task.^a

	HR (beats/min)	TP (ms²)	HF (ms²)	LF (ms²)	nuHF	nuLF	LF/HF
Volition	74.0 ± 1.6**	313.0 ± 52.2	98.0 ± 21.4**	197.0 ± 35.7	31.2 ± 2.3**	68.8 ± 2.3**	2.8 ± 0.3**
Baseline	71.0 ± 1.5	338.0 ± 51.0	154.0 ± 25.2	167.0 ± 31.2	47.4 ± 3.4	52.7 ± 3.4	1.5 ± 0.2

Abbreviations: HRV, heart rate variability; HR, heart rate; TP, total power; HF, high frequency; LF, low frequency; nuHF, normalized high frequency; nuLF, normalized low frequency.

Data are presented as mean ± standard error of the mean.

P < .05 versus baseline (paired t test).

P < .01 versus baseline (paired t test).

Figure 3 represents the change of RR intervals with time in HR_down and HR^up task. The changes existed in all subjects, so a single subject was selected as a typical example. It showed that changes of heartbeat were delayed about 40 seconds after the start of volitional control.

Brain Activity in Volitional Control Task

Wavelet packet energy of EEG frequency bands of alpha and beta during the slowing and quickening exercises are illustrated in Figures 4 and 5, respectively. Compared with the baseline, a significant increase was found in alpha power at FC3, FC4, FCz, C3, C4, Cz, P3, P4, Pz, and T7 electrodes (Figure 4A; P < .05), and a significant decrease in beta power at FCz and C4 electrodes in HR_down task (Figure 4B; P < .05). In the HR^up task, a significant increase was found in alpha power at P4, O1, O2, and Oz electrodes (Figure 5A; P < .05), and an increase in beta power at FC4, C4, Cz, P4, Pz, O1, O2, and Oz electrodes was found (Figure 5b; P < .05). The prefrontal areas were activated independent of the sound in the HR_down task, and same phenomenon was observed in the precentral, central, parietal, and occipital areas in the HR^up task.

Information Flow Between Brain Regions

During volitional control the DTFdiff values from Pz to FC3, C3, C4, T7, and P3 and the values from P4 to C4 were all positive in the HR_down task. Compared with the baseline, these DTFdiff values significantly increased at alpha band. At beta band, the DTFdiff values from Pz to P3 and C4 significantly increased. In the HR^up task, the DTFdiff values from Pz to FC3, FC4, C3, C4, T7, T8, P3 and the values from P4 to FC3, FC4, C3, C4 showed significantly increased alpha compared with baseline. The DTFdiff values from Pz to FC3, C4, T7 and the values from P4 to FC3, FC4 significantly increased at beta band. The information flow from parietal regions to central and precentral areas was dominant when the heartbeat was controlled to slow down or quicken (Figures 6 and 7; P < .05).

Information Flow Between Brain and Heart

When the heartbeat was controlled to slow down, the DTFdiff value to the RR interval was positive in beta at the C4 location, whereas it was negative at baseline. When it was controlled to quicken, the DTFdiff values from FC4 and C4 locations to RR interval were positive in beta, and negative at baseline (Figure 8; P < .05). At baseline dominant information flowed from heart to brain. When the heartbeat was controlled to slow down or quicken, the directional information flow from the central areas or frontal–central areas of brain to heart were dominant.

Discussion

Many diseases are related to arrhythmia, and some antiarrhythmic medicines can themselves paradoxically lead to life-threatening rhythm disorders and increase the mortality. There has been, therefore, an increasing shift toward nonpharmacological therapies for cardiac arrhythmias.³⁸ The ANS has a significant effect on the heart, and nonpharmacological methods related to the ANS have improved arrhythmia. In this study, volitional experiments were designed to change heart rate. HR was changed by volition, and it significantly increased and decreased, respectively, in the HR^up and HR_down tasks (Tables 1 and 2). The range of HR variation affected by volition was approximately more than 2 beats. Although this performance was not so effective, it proved that heartbeat can be partly controlled by volition. The study may provide a potential and convenient method to improve diseases related to HR when the patients are trained and practice volition.

It is known that sympathetic and parasympathetic activities affect HR. In this study, nuLF and LF/HF significantly decreased, and nuHF significantly increased, when HR was controlled to slow down (Table 1). It indicated that parasympathetic activity became strong and sympathetic activity weak in the HR_down task. It was the change of the autonomic nervous activity that decreased HR. Similarly, strong sympathetic and weak parasympathetic activity induced the increase of HR in the HR^up task (Table 2). Our results showed that parasympathetic and sympathetic activities were changed during slowing or quickening induced by volition. Our method may also be valuable to drive the ANS in other related studies. In this study, cardiac autonomic activity was obtained from power spectral analysis of RR intervals. This method is widely used to estimate sympathetic and parasympathetic tone. Several approaches have been used to assess a time-varying power spectrum as the reference mentioned.³⁹ The time-dependent power spectrum of HR fluctuations is a valuable direction and will be further studied in the future.

Giving a command to move a hand, is initiated at some special cortical positions. The electrical activity of neurons in the positions will be changed while the command is issued. Alpha and beta are the dominant rhythms in teenagers and adults. When the alpha band increases and the beta band is subdued, synchronized activity of brain tissue is significant. Beta waves indicate excitability of the brain and occur in individuals who exert specific mental effort.^40,41 The results of the wavelet packet analysis showed that alpha waves increased and beta waves decreased in the HR_down task. In the HR^up task, both alpha and beta waves increased (Figures 4 and 5). The alpha and beta waves can be increased during acupuncture or in a high arousal state.^42,43 Results revealed that synchronization of the brain increased, and the prefrontal areas were markedly changed when the heartbeat was controlled to slow down. When the heartbeat was quickened, a more excited state at the precentral, central, parietal, and occipital areas was observed.

Although some researchers have suggested that the frontal and postcentral areas are associated with autonomic activity, the mechanism of the scalp neuron activity and the interactions between anterior-posterior brain regions participating in sympathetic and parasympathetic activity remains unclear. Anatomical studies indicated high connectivity between frontal and posterior cortices.⁴⁴ In this study, DTF was used to investigate the functional relationship between different brain regions. Results showed that the information flow from parietal areas to precentral areas increased. The parietal areas were the “source” locations of information outflow in volitional control task (Figures 6 and 7). Our findings suggest that during the volitional control of heartbeat, the dominant direction was from parietal areas to sensorimotor cortex, which indicate that the parietal areas played an important role to trigger the sensorimotor cortex and integrate information.

It has been found that autonomic dysfunction occurred in the presence of hemispheric lesions involving sensory pathways from the cortex to the internal capsule and insular.⁴⁵ So far, many studies have investigated the brain regions that may influence cardiovascular functions. The feedback from heart to brain has also been studied.^14,17,36,46 However, functional coupling patterns of brain and heart activity have not been adequately studied. To better investigate the mechanism and functional relevance of brain and heart system, DTF based on MVAR was used to study the directional coupling from brain to heart. The results demonstrated that the information flow from the central and precentral locations to the heart was dominant, and it mainly showed active regulation of brain-heart. Previous studies have found that C4 and Cz electrodes corresponded to somatosensory areas and had partial function as a vegetative nerve center.^47,48 Heartbeat perception experiments also indicated that the sensorimotor area was the main activity area of brain in the heartbeat perception process.⁴⁹ The coherence of precentral areas was correlated with autonomic nervous activity during meditation.²² Our results further indicated that directional information flow propagated from somatosensory areas to heart. Our observations should improve understanding of cortical areas in modulating the ANS and comprehension of the pathways from pallium to heart.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The work was funded by the National Natural Science Foundation of China (No. 31170893 and No. 81271659).

Full-color figures are available online at

References

Middlekauff

. Acupuncture in the treatment of heart failure. Cardiol Rev. 2004;12:171-173.

Oppenheimer

. Cerebrogenic cardiac arrhythmias: cortical lateralization and clinical significance. Clin Auton Res. 2006;16:6-11.

Thayer

Brosschot

. Psychosomatics and psychopathology: looking up and down from the brain. Psychoneuroendocrinology. 2005;30:1050-1058.

Hilz

Devinsky

Szczepanska

Borod

Marthol

Tutaj

. Right ventromedial prefrontal lesions result in paradoxical cardiovascular activation with emotional stimuli. Brain. 2006;129(pt 12):3343-3355.

Inagaki

Kuwahara

Tsubone

. Effects of psychological stress on autonomic control of heart in rats. Exp Anim. 2004;53:373-378.

Marci

Glick

Loh

Dougherty

. Autonomic and prefrontal cortex responses to autobiographical recall of emotions. Cogn Affect Behav Neurosci. 2007;7:243-250.

McDougall

Widdop

Lawrence

. Central autonomic integration of psychological stressors: focus on cardiovascular modulation. Auton Neurosci. 2005;123:1-11.

Sebastiani

Simoni

Gemignani

Ghelarducci

Santarcangelo

. Autonomic and EEG correlates of emotional imagery in subjects with different hypnotic susceptibility. Brain Res Bull. 2003;60:151-160.

Nakao

Yano

Nomura

Kuboki

. Blood pressure-lowering effects of biofeedback treatment in hypertension: a meta-analysis of randomized controlled trials. Hypertens Res. 2003;26:37-45.

10.

Schwartz

. Voluntary control of human cardiovascular integration and differentiation through feedback and reward. Science. 1972;175:90-93.

11.

Jayasinghe

. Yoga in cardiac health (a review). Eur J Cardiovasc Prev Rehabil. 2004;11:369-375.

12.

Shannahoff-Khalsa

Sramek

Kennel

Jamieson

. Hemodynamic observations on a yogic breathing technique claimed to help eliminate and prevent heart attacks: a pilot study. J Altern Complement Med. 2004;10:757-766.

13.

Skoglund

Jansson

. Qigong reduces stress in computer operators. Complement Ther Clin Pract. 2007;13:78-84.

14.

Kimmerly

O’Leary

Menon

Gati

Shoemaker

. Cortical regions associated with autonomic cardiovascular regulation during lower body negative pressure in humans. J Physiol. 2005;569(pt 1):331-345.

15.

Verberne

Owens

. Cortical modulation of the cardiovascular system. Prog Neurobiol. 1998;154:149-168.

16.

Critchley

Mathias

Dolan

. Neural activity in the human brain relating to uncertainty and arousal during anticipation. Neuron. 2001;29:537-545.

17.

Critchley

Corfield

Chandler

Mathias

Dolan

. Cerebral correlates of autonomic cardiovascular arousal: a functional neuroimaging investigation in humans. J Physiol. 2000;523:259-270.

18.

Lumbers

Yan

. A method for determining baroreflex-mediated sympathetic and parasympathetic control of the heart in pregnant and non-pregnant sheep. J Physiol. 1999;515:555-566.

19.

Malliani

Lombardi

Pagani

Cerutti

. Power spectral analysis of cardiovascular variability in patients at risk for sudden cardiac death. J Cardiovasc Electrophysiol. 1994;5:274-286.

20.

Mateo

Laguna

. Improved heart rate variability signal analysis from the beat occurrence times according to the IPFM model. IEEE Trans Biomed Eng. 2000;47:985-996.

21.

Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation. 1996;93:1043-1065.

22.

Murata

Takahashi

Hamada

. Individual trait anxiety levels charactering the properties of zen meditation. Neuropsychobiology. 2004;50:189-194.

23.

Stam

Van Dijk

. Synchronization likelihood: an unbiased measure of generalized synchronization in multivariate data sets. Physica D. 2002;163:236-251.

24.

Urbano

Babiloni

Onorati

Babiloni

. Dynamic functional coupling of high resolution EEG potentials related to unilateral internally triggered one-digit movements. Electroencephalogr Clin Neurophysiol. 1998;106:477-487.

25.

Kaminski

Blinowska

. A new method of the description of the information flow in the brain structures. Biol Cybern. 1991;65:203-210.

26.

Pereda

Quiroga

Bhattacharya

. Nonlinear multivariate analysis of neurophysiological signals. Prog Neurobiol. 2005;77:1-37.

27.

Mima

Matsuoka

Hallett

. Information flow from the sensorimotor cortex to muscle in humans. Clin Neurophysiol. 2001;112:122-126.

28.

Vecchio

Del Percio

Marzano

. Functional cortico-muscular coupling during upright standing in athletes and nonathletes: a coherence electroencephalographic-electromyographic study. Behav Neurosci. 2008;122:917-927.

29.

McCraty

. Influence of cardiac afferent input on heart-brain synchronization and cognitive performance. Int J Psychophysiol. 2002;45:72-73.

30.

Brunner

Hemingway

Walker

. Adrenocortical, autonomic, and inflammatory causes of the metabolic syndrome: nested case-control study. Circulation. 2002;106:2659-2665.

31.

Salomé

Ngampramuan

Nalivaiko

. Intra-amygdala injection of GABAA agonist, muscimol, reduces tachycardia and modifies cardiac sympatho-vagal balance during restraint stress in rats. Neuroscience. 2007;148:335-341.

32.

Stein

Bosner

Kleiger

Conger

. Heart rate variability: a measure of cardiac autonomic tone. Am Heart J. 1994;127:1376-1381.

33.

Hilz

Dütsch

. Quantitative studies of autonomic function. Muscle Nerve. 2002;33:6-20.

34.

Malliani

Pagani

Lombardi

Cerutti

. Cardiovascular neural regulation explored in the frequency domain. Circulation. 1991;84:482-492.

35.

Hwa

Ferree

. Scaling properties of fluctuations in the human electroencephalogram. Phys Rev E Stat Nonlin Soft Matter Phys. 2002;66(2 pt 1):021901.

36.

Gray

Taggart

Sutton

. A cortical potential reflecting cardiac function. Proc Natl Acad Sci U S A. 2007;104:6818-6823.

37.

De Gennaro

Vecchio

Ferrara

Curcio

Rossini

Babiloni

. Antero-posterior functional coupling at sleep onset: changes as a function of increased sleep pressure. Brain Res Bull. 2005;65:133-140.

38.

Nattel

. New ideas about atrial fibrillation 50 years on. Nature. 2002;415:219-226.

39.

Toledo

Gurevitz

Hod

Eldar

Akselrod

. Wavelet analysis of instantaneous heart rate: a study of autonomic control during thrombolysis. Am J Physiol Regul Integr Comp Physiol. 2003;284:R1079-R1091.

40.

Gianaros

Van Der Veen

Jennings

. Regional cerebral blood flow correlates with heart period and high-frequency heart period variability during working-memory tasks: implications for the cortical and subcortical regulation of cardiac autonomic activity. Psychophysiology. 2004;41:521-530.

41.

Liu

Wen

Yao

. The Brain and Nonlinear Dynamics. Beijing, China: Science Press; 2006.

42.

Morinushi

Masumoto

Kawasaki

Takigawa

. Effect on electroencephalogram of chewing flavored gum. Psychiatry Clin Neurosci. 2000;54:645-651.

43.

Sakai

Hori

Umeno

Kitabayashi

Ono

Nishijo

. Specific acupuncture sensation correlates with EEGs and autonomic changes in human subjects. Auton Neurosci. 2007;133:158-169.

44.

Goldman-Rakic

. Topography of cognition: parallel distributed networks in primate association cortex. Annu Rev Neurosci. 1998;11:137-156.

45.

Diserens

Vuadens

Miche

. Acute autonomic dysfunction contralateral to acute strokes: a prospective study of 100 consecutive cases. Eur J Neurol. 2006;13:1245-1250.

46.

Tanida

Sakatani

Takano

Tagai

. Relation between asymmetry of prefrontal cortex activities and the autonomic nervous system during a mental arithmetic task: near infrared spectroscopy study. Neurosci Lett. 2004;369:69-74.

47.

Ginter

Jr Blinowska

Kamiński

Durkap

. Phase and amplitude analysis in time-frequency space application to voluntary finger movement. J Neurosci Methods. 2001;110:113-124.

48.

Garey

. Brodmann’s Localisation in the Cerebral Cortex. London, UK: Smith-Gordon; 1994.

49.

Yuan

Yan

. Effect of heartbeat perception on heartbeat evoked potential waves. Neurosci Bull. 2007;23:357-362.