Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Postural control adaptation and habituation during vibratory proprioceptive stimulation: an HD-EEG investigation of cortical recruitment and kinematics

Postural control adaptation and habituation during vibratory proprioceptive stimulation: an... Postural control adaptation and habituation during vibratory proprioceptive stimulation: an HD-EEG investigation of cortical recruitment and kinematics Fabio Barollo, Rún Friðriksdóttir, Kyle J. Edmunds, Gunnar H. Karlsson, Halldór Á. Svansson, Mahmoud Hassan, Antonio Fratini, Hannes Petersen and Paolo Gargiulo equilibrium condition of upright stance [1], [2]. Collectively Abstract— The objective of the present work is to measure defined as ‘postural control’ [3], this process is dynamically postural kinematics and power spectral variation from HD-EEG mediated by regulatory feedback elicited from somatosensory, to assess changes in cortical activity during adaptation and vestibular, and visual systems [4]. Exogenous disruption or habituation to postural perturbation. To evoke proprioceptive nonspecific stimulation of these systems can induce postural postural perturbation, vibratory stimulation at 85 Hz was applied sway [5], [6] – the magnitude and latency of which characterize to the calf muscles of 33 subjects over four 75-second stimulation the kinematics of postural control, which can be assessed by periods. Stimulation was performed according to a pseudorandom binary sequence. Vibratory impulses were synchronized to high- changes in force and torque actuated at the support surface of density electroencephalography (HD-EEG, 256 channels). the feet, altogether known as posturography [7], [8], [9]. Changes in absolute spectral power (ASP) were analyzed over four Postural control measurement is often employed under frequency bands (Δ: 0.5-3.5 Hz; θ: 3.5-7.5 Hz; α: 7.5-12.5 Hz; β: conditional balance perturbation, which is typically achieved 12.5-30 Hz). A force platform recorded torque actuated by the feet, via visual disturbance or proprioceptive stimulation [10]. and normalized sway path length (SPL) was computed as a Extant research in this regard cites postural control as a construct for postural performance during each period. SPL values indicated improvement in postural performance over the fundamental ‘learned’ motor skill, whose function and trial periods. Significant variation in absolute power values (ASP) efficiency can be systematically improved with routine postural was found in assessing postural adaptation: an increase in θ band tasks [9] or directed training [11]. From these studies, two ASP in the frontal-central region for closed-eyes trials, an increase dimensions of postural learning have been posited: in θ and β band ASP in the parietal region for open-eyes trials. In ‘adaptation’, defined as transient improvements in motor habituation, no significant variations in ASP were observed during response to upright balance perturbation [12], and ‘habituation’, closed-eyes trials, whereas an increase in θ, α, and β band ASP was observed with open eyes. Furthermore, open-eyed trials generally conversely defined by a gradual decrease in response to yielded a greater number of significant ASP differences across all repeated perturbation [13]. bands during both adaptation and habituation, suggesting that following cortical activity during postural perturbation may be While the response of sensorimotor systems during postural up-regulated with the availability of visual feedback. These results adaptation and habituation has been well-established in altogether provide deeper insight into pathological postural literature [14], [15] interrogating the commensurate role of the control failure by exploring the dynamic changes in both cortical activity and postural kinematics during adaptation and cerebral cortex or subcortical central nervous system (CNS) habituation to proprioceptive postural perturbation. structures is a comparatively recent subject of research. In this regard, literature has extended previous knowledge on Index Terms— Balance, cerebral cortex, HD-EEG, kinematics, subcortical balance maintenance [16], [17] to consider the postural control, power spectral density potential governing role of supratentorial information processing in the cerebral cortex [18]–[20]. The neuroimaging I. INTRODUCTION method of electroencephalography (EEG) has been cited for its UMAN posture is a complex and naturally unstable high temporal resolution in measuring cortical activity [6], [21], physiological process that requires the continuous [22]. In postural control research, balance perturbation has integration of compensatory mechanisms to maintain an Manuscript received MM DD, YY. Date of publication MM DD, YY. Fabio Barollo is with Reykjavik University and the Department of This research was supported jointly by the Institute for Biomedical and Biomedical Engineering at Aston University, Birmingham, UK. Neural Engineering at Reykjavík University, the Department of Anatomy at Antonio Fratini is with the Department of Biomedical Engineering at Aston the University of Iceland, and the Icelandic National Hospital (Landspítali University, Birmingham, UK. Scientific Fund) with additional funding support from the Rannís Icelandic M. Hassan is with LTSI- Université de Rennes, Rennes, France. Research Fund. Asterisk indicates corresponding author. H. Petersen is with The Department of Anatomy, University of Iceland, R. Friðriksdóttir, K. J. Edmunds, G.H. Karlsson, H.Á. Svansson and *P. Reykjavík, Iceland and with Akureyri Hospital, 600 Akureyri, Iceland Gargiulo are with Reykjavík University, Reykjavík, Iceland (Fabio Barollo and Rún Friðriksdóttir are co-first authors.) (correspondence e-mail: paolo@ru.is). 2 Fig. 1. Experimental set-up. revealed scalp-level activity changes in frontal-central and frontal-parietal cortical regions, specifically within α (7.5–12.5 II. MATERIALS AND METHODS Hz) and θ (3.5–7.5 Hz) frequency ranges [20], [23]. Additional A. Experimental setup EEG studies report bursts of γ activity (30–80 Hz) during voluntarily anterior-posterior movements [13]. EEG activity in As noted, the present experimental setup aimed at this regard is reported as changes in evoked time-domain event- integrating posturography measurement with the assessment of related potentials (ERP) or as perturbation-evoked responses cortical activity by 256-channel HD-EEG during vibratory (PERs), e.g. N1 amplitudes and Contingent Negative Variations proprioceptive stimulation. The study was reviewed and (CNV) [3], [6], [19], [24]. approved by the National Bioethics Committee (Vísindasiðanefnd - reference number: VSN-063) and the While recent evidence for the critical role of the cerebral measurements were performed at the Icelandic Center for cortex in governing postural adaptation and habituation has Neurophysiology at Reykjavik University. Thirty-three healthy been reported [10], to our knowledge, cortical activity volunteers (10 females and 23 males, aged 21 to 52) assessment during balance perturbation has never been participated in the study. These subjects had no history of synchronized with kinematic posturography measurements. In vertigo, central nervous disease, or lower extremity injury, and addition, no postural control studies report the use of ‘high- none of the subjects had consumed alcohol within a 24-hour density’, 256-channel EEG (HD-EEG) – a methodology with period prior to their measurement. Fig. 1 illustrates the overall superior spatial resolution to more conventional 32- or 64- experimental set-up for the present work, as presented at the XV channel systems. Furthermore, power spectral variation Mediterranean Conference on Medical and Biological analysis from EEG data remains underreported in postural Engineering and Computing – MEDICON 2019 [26]. control research, despite its being a conventional method for Participants were instructed to maintain an upright stance EEG signal analysis with proven utility in cognitive and motor during exogenous balance perturbation, evoked by the task studies [20], [25]. The present study aims to extend current simultaneous stimulation of vibrators fastened tightly by elastic research in this regard with the synchronized assessment of straps around the widest point of each calf. postural kinematics with power spectral variation analyses from The vibrators were designed using revolving DC-motors HD-EEG to quantify changes in cortical activity during equipped with a 3.5 gram eccentric weight, which was adaptation and habituation to postural perturbations using contained in a cylindrical casing approximately 6 cm in length vibratory proprioceptive stimulation. and 1 cm in diameter. Each stimulation was set to deliver 3 vibrations of 0.1 cm in amplitude, at a frequency of 85 Hz. III. DATA ANALYSIS Stimulation was applied according to a pseudorandom binary A. Assessing postural performance sequence schedule (PRBS), where each shift had a random duration of 1 to 6.6 seconds. Force platform data for both OE and CE datasets were B. Posturography measurement analogously segmented into five recording periods (QS, P1, P2, P3, and P4) in order to facilitate the synchronization of postural In general, maintaining a normative eased upright stance requires the symmetric distribution of body weight; when challenged, the resultant anterior-posterior or bilateral compensatory motion can be captured using a force platform to record changes in the body’s center of pressure [27], [28]. For the present work, this assessment was achieved using a customized platform system developed at the Department of Solid Mechanics, Lund Institute of Technology in Sweden [3]. Anterior-posterior (ant-post) and lateral (Lat) forces actuated by the feet were recorded at six degrees of freedom with an accuracy of 0.5 N; these data were sampled at 50 Hz by a custom-made program, Postcon™, on a computer equipped with an analogue-to-digital converter. Participants were instructed to stand on a pressure plate with their arms downwardly relaxed and their feet positioned at an angle of approximately 30 degrees, open to the front, with their heels approximately three centimeters apart. Participants were asked to focus on a fixed marker point in front of them, at about 150 Fig. 2. Statokinesigram showing an example subject’s body sway. Left: no cm distance. stimulation applied. Right: randomized vibratory stimulation applied on the calves to disrupt upright stance. Anterior- posterior and lateral displacements are expressed in cm. C. HD-EEG data acquisitions sway data with any evoked changes in cortical activity. The normalized Sway Path Length (SPL) was computed to describe HD-EEG data were acquired using a 256-channel Ag/AgCl the overall postural performance of the subjects during each wet-electrode cap connected in bipolar configuration to four period. cascaded 64-channel amplifiers; data collection employed a standardized 10-20 system montage with EEGO software (ANT The center of pressure (CoP) trajectory coincides with the neuro, Enschede Netherlands). An additional infra-orbital vertical projection on platform plane of the subject’s center of electrode was used to identify any obfuscating mass and it is widely used in human posture studies to assess electrooculographic (EoG) signals, and EEG data were body sway [29], [30]. Torque values were extracted from the continuously recorded at a sampling frequency of 1024 Hz. To collected force platform data and used to derive ant-post and lat synchronize EEG acquisition with posturography data, a displacement during each trial. A graphical representation of custom trigger signal box was built to rectify each vibratory the CoP trajectory (‘statokinesigram’) was then obtained by stimuli as a 5V TTL timing signal sent to the master amplifier. plotting the displacement along the ant-post and lat axes over This trigger system allowed for the generation of vibratory time, see Fig. 2. Sway Path Length is one of several on/off event timestamps at <1 ms latency during EEG posturographic parameters that can be extracted from a postural recordings. As our previous research has identified changes in platform and is one of the most commonly used [31]. cortical activity according to the availability of visual feedback, Nevertheless, previous research has established the sensitivity two measurement trials were performed for each subject, of these indices to different anthropometric characteristics – beginning with open eyes (OE) and followed by closed eyes particularly height and weight, where taller or heavier people (CE). An initial quiet stance (QS) baseline phase (30 seconds) appear to be more unstable [32]. As such, SPL values were preceded each stimulation phase (300 seconds), resulting in a normalized to each subject’s height and weight to estimate the total duration of 330 seconds for each recording. Each CoP trajectory on the platform according to the following stimulation phase was further subdivided into four 75-second formula: recording periods: P1, P2, P3, and P4. 𝑔𝑟𝑆𝑡𝑙𝑜𝑎𝑚𝑎𝑏𝑖 (𝑡 ) = ( ) ( ) 0.56∙ℎ ∙ 𝑚 ∙9.81 𝑘 𝑘 TABLE I Normalized Sway Path Length (SPL), means, and standard deviations by experimental epoch and OE/CE condition. QS P1 P2 P3 P4 Eyes Open Closed Open Closed Open Closed Open Closed Open Closed Mean 0.3038 0.4408 1.3624 1.8584 1.2190 1.7276 1.2312 1.7229 1.2574 1.6655 SD 0.0821 0.1392 0.2810 0.5130 0.2389 0.4447 0.2285 0.4204 0.2422 0.4285 Where 𝑘 refers to the individual participant, 0.56∙ℎ is an heteroscedastic t-tests, which yielded topological p-value maps estimate of the center of mass’s (CoM) radial distance to the for each EEG waveform (with p<0.05 the threshold for platform [33] and 𝑚 ∙9.81 is the weight applied to the significance). False discovery rate (FDR) and Bonferroni participant’s CoM. significance correction methods were employed and compared to address the statistical problem of multiple comparisons. B. HD-EEG data preprocessing Channels that resisted these multiple comparison correction methods were specifically marked in significance topologies. Raw HD-EEG data were pre-processed using the EEGLAB Toolbox [34], first with a band-pass filter set between 0.5–80 IV. RESULTS Hz, followed by a notch filter (49.5–50.5Hz) to remove AC A. SPL data distribution to assess adaptation and habituation power line interference from each period. Three approaches to periods artifact rejection were performed: channel interpolation, automatic continuous artifact rejection, and principle When describing time-varying phenomena, a distinction component analysis (PCA). All channels were re-referenced to must be made between true adaptation resulting from control a common average. dynamic alterations and other time-dependent phenomena, such as fatigue or biomechanical alterations resultant from changes To investigate variations in cortical activity between in balance conditions [3]. Because of this, we aimed to use SPL periods, absolute spectral power (ASP) values were obtained results to observe whether postural performance decreased over using fast Fourier transformation (FFT) analysis at a resolution any of the experimental periods; with both OE and CE of 0.977 Hz with a 10% Hanning window. This analysis was conditions, minimum postural sway has been reported to occur performed for four frequency bands: Δ (0.5-3.5 Hz), θ (3.5-7.5 at around 150-200 seconds following incident stimuli [3], [10]. Hz), α (7.5-12.5 Hz) and β (12.5-30 Hz). This analysis Table I shows the means and standard deviations of the generated ASP values for each frequency band and period, normalized SPL values for the present cohort, across each of which were extracted and exported for statistical analyses using the experimental periods and conditions. Fig. 3 represents the a customized Matlab GUI (MathWorks, Inc., Natick, 158 SPL data distribution with box plots to highlight median, mean Massachusetts, USA). From this analysis, topological maps and outlier values for each period. From these results, SPL did illustrating OE and CE difference spectra were extracted for increase in Period 4 (P4) for OE trials, but remained each EEG frequency band. In addition, significant changes in comparatively stable over all four periods for CE trials. To mean ASP at each electrode were assessed using paired Fig. 3.Normalized Sway Path Length. Left: Open Eyes; Right: Closed Eyes. avoid any potential obfuscation from fatigue or other time- Adaptation: dependent variations, we therefore used the difference between Period 3 (P3) and Period 1 (P1) ASP values in our assessment As previously mentioned, postural adaptation is shown of habituation. using normalized ASP differences between P1 and QS, and Figs. 4 and 5 depict cortical maps of channels showing significant changes in CE and OE conditions, respectively. B. Absolute power spectra variation from HD-EEG These results show an increase in ASP for both conditions, with many more significant regions present from OE trials. Figs. 4–6 show the results from statistical analyses Generally, ASP in the θ band increased in adaptation – performed across every channel. Each of the different colored particularly in the frontal-central region (p<0.05, FDR regions defines a topological map of statistical significance corrected) during CE and the parietal region (p<0.05, FDR (p<0.05) corrected for multiple comparisons in accordance with corrected) during OE, where ten electrodes passed the the following conditions: uncorrected paired t-test (yellow), Bonferroni correction test. During OE trials, higher frequency FDR correction (orange), and Bonferroni correction (red). bands (α and β) show significant activity as well. In particular, Channels that remained significant following FDR correction Fig. 5 shows increased activity in the parietal-occipital region are highlighted with a black ‘X’, while electrodes that resisted in the β band. Bonferroni correction are indicated by a blue ‘*’. White regions are considered non-significant (p>0.05). In the second row, Habituation: topographical maps of changes in ASP are shown solely for areas that presented a statistically significant difference Postural habituation is shown using normalized ASP between the examined periods. differences between P3 and P1. Fig. 6 depicts the cortical mapping of grand mean changes in ASP solely in channels shown to have a significant difference between periods Fig. 4. Adaptation CE. Top row: statistical analysis results. Each colored region identifies topological areas that achieved statistical significance (p<0.05) given the following conditions: uncorrected paired t-test (yellow), FDR correction (orange), and Bonferroni correction (red). Individual electrodes that resisted FDR correction are indicated by a black ‘X’. Bottom row: Topographies highlighting the average changes in ASP (between P1 and QS periods) over the whole cohort. Fig. 5. Adaptation OE. Top row: statistical analysis results. Each colored region identifies topological areas that achieved statisti cal significance (p<0.05) given the following conditions: uncorrected paired t-test (yellow), FDR correction (orange), and Bonferroni correction (red). Individual electrodes that resisted FDR correction are indicated by a black ‘X’, while electrodes that resisted Bonferroni correction are indicated by a blue ‘*’. Bottom row: Topographies highlighting the average changes in ASP (between P1 and QS periods) over the whole cohort. mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz 6 Fig. 6. Habituation OE. Top row: statistical analysis results. Each colored region identifies topological areas that achieved statistical significance (p<0.05) given the following conditions: uncorrected paired t-test (yellow), FDR correction (orange), and Bonferroni correction (red). Individual electrodes that resisted FDR correction are indicated by a black ‘X’. Bottom row: Topographies highlighting the average changes in ASP (between P3 and P1 periods) over the whole cohort. (p<0.05). White areas are again considered non-significant. The post direction. In all of these comparisons, no meaningful overall results show an increase of ASP in P3 compared to P1 correlation was found, suggesting the future importance of for OE trials, whereas during the CE periods, no significant considering alternative metrics for postural performance, such changes occurred. OE results specifically indicate increased as approximate entropy or multiscale entropy, which has shown activity in the θ band (temporal region), α band (parietal promise in linking modifications in neural involvement to region), and β band (frontal region). responses to vibration [35]. V. DISCUSSION Significant ASP differences were found across the entirety of the cortex in α and θ bands, with the exception of the prefrontal Power spectral variation analysis from EEG data remains area which yielded minimal significance in the α band. Power underreported in postural control literature. The present study in the θ band has been shown to increase in adaptation – aimed to extend current research with the synchronized particularly in the frontal-central region during CE trials and the assessment of postural kinematics with power spectral variation parietal area during OE trials [20], [23], [36]. Here, our results analyses from HD-EEG to quantify changes in cortical activity are in accordance with the recent work by Solis-Escalante et al., during adaptation and habituation during a postural control task who showed significant midline ASP differences in both CE using vibratory proprioceptive stimulation. and OE conditions as evidenced by the differential modulations Postural sway was recorded to obtain normalized SPL of α and low-γ rhythms [37]. This altogether suggests that values over five experimental periods, which were then utilized central region ASP increases during high-demand postural to classify postural adaptation and postural habituation as correction, such as balance maintenance without allowing differences in ASP between specific recording periods. In this corrective foot placement, as performed in the present study. regard, adaptation was defined as the difference between P1 and QS periods, in accordance with extant postural perturbation Furthermore, it has been suggested that increase in θ activity literature [3], [10], [12]. However, normalized SPL was used to in the frontal-central regions is involved in error detection and define postural habituation, where P1 ASP values were processing of postural stability during balance control [22], subtracted from the period with the lowest mean SPL to avoid [38]. As such, the θ band ASP differences shown here may the influence of time-dependent variations such as fatigue. signify the planning of corrective steps and/or the analysis of falling consequences, as indicated by our previous work on There are two main limitations to note from this cortical functional dynamics during postural control. Relatedly, methodology. Firstly, while it remains possible that the onset of significant ASP differences in the α band may reflect an habituation or fatigue may differ subject-to-subject, previous inhibition of error detection within the cingulate cortex due to studies have shown that ant-post and lat torque variations habituation [10]. during OE and CE trials reach a consistent minimum around 150-200 seconds following incident stimuli, corresponding The present results indicate that OE trials reflect a greater with P3 in the present work [3], [10]. Nevertheless, further number of significant differences in ASP across all bands investigation into quantitative indicators for the measurement during both adaptation and habituation. This suggests that of cortical habituation is recommended. Another following both acute and prolonged proprioceptive methodological limitation arises from demonstrating the perturbation, cortical activity may be up-regulated with the relationship between ASP and SPL. In this regard, we availability of visual feedback. These results generally support investigated whether there was any linear correlation between our prior hypothesis that the visual recognition of instability ASP and SPL or other metrics derived from CoP trajectory, may play a critical role in governing cortical processes requisite such as the root-mean-square (RMS) of excursion in the ant- mV /Hz mV /Hz mV /Hz mV /Hz 7 stabilization of human posture after a sudden disturbance: influence for postural control [10]. This hypothesis resulted from of rate and amplitude of displacement,” Exp. Brain Res., 1984. previous work demonstrating that all-type visual impairment is [6] K. L. Goh, S. Morris, W. L. Lee, A. Ring, and T. Tan, “Postural and associated with an increased risk for injurious falling [39], [40]. cortical responses following visual occlusion in standing and sitting tasks,” Exp. Brain Res., 2017. However, in addition to the impact of vision, it is also important [7] L. A. Cohen, “Role of eye and neck proprioceptive mechanisms in to note that postural control may be driven by brain network body orientation and motor coordination,” J. Neurophysiol., 1961. interactions rather than isolated changes in cortical activity at [8] M. H. Hu and M. H. Woollacott, “Multisensory training of standing balance in older adults: I. Postural stability and one-leg stance specific regions. Our present results support the potential balance,” Journals Gerontol., 1994. importance of network dynamics, as significant differences in [9] T. S. Galvão, E. S. Magalhães Júnior, M. A. Orsini Neves, and A. ASP were concurrently measured across many different cortical de Sá Ferreira, “Lower-limb muscle strength, static and dynamic regions. In this context, we believe that further work postural stabilities, risk of falling and fear of falling in polio survivors and healthy subjects,” Physiotherapy Theory and investigating the reconfiguration of cortical networks during Practice, 2018. adaptation and habituation could reveal new insights about how [10] K. J. Edmunds et al., “Cortical recruitment and functional dynamics functionally coordinated brain activity may dictate postural in postural control adaptation and habituation during vibratory proprioceptive stimulation,” J. Neural Eng., 2019. control. Such a model could include source-space or [11] A. S. Pollock, B. R. Durward, P. J. Rowe, and J. P. Paul, “What is connectivity-based analyses, as we have previously illustrated balance?,” Clin. Rehabil., vol. 14, no. 4, pp. 402–406, 2000. using lower density EEG [10]. Furthermore, since the examined [12] T. D. J. Welch and L. H. Ting, “Mechanisms of motor adaptation in reactive balance control,” PLoS One, 2014. measures of postural performance (SPL and RMS) were not [13] J. C. Eccles, “Learning in the motor system,” Prog. Brain Res., vol. able to highlight a significant correlation between kinematics 64, pp. 3–18, Jan. 1986. and cortical recruitment, a deeper investigation based on [14] D. A. E. Bolton and J. E. Misiaszek, “Compensatory balance reactions during forward and backward walking on a treadmill,” different postural parameters would be a fruitful area for further Gait Posture, 2012. work. [15] C. F. Honeycutt and T. Richard Nichols, “Disruption of cutaneous feedback alters magnitude but not direction of muscle responses to postural perturbations in the decerebrate cat,” Exp. Brain Res., 2010. VI. CONCLUSIONS [16] R. Magnus, “Some results of studies in the physiology of posture,” Dynamic posturography is an established method for Lancet, 1926. evaluating postural control. By delivering controlled disruption [17] C. S. Sherrington, “Flexion‐reflex of the limb, crossed extension‐ reflex, and reflex stepping and standing,” J. Physiol., 1910. in the form of calf vibration, we can examine central nervous [18] A. Mierau, B. Pester, T. Hülsdünker, K. Schiecke, H. K. Strüder, system (CNS) processing by associating body inertia and and H. Witte, “Cortical Correlates of Human Balance Control,” changes in upright stance [41]. This notion is true for simple Brain Topogr., 2017. [19] J. V. Jacobs and F. B. Horak, “Cortical control of postural upright stance but does not apply to complex postural tasks or responses,” in Journal of Neural Transmission, 2007. pathological conditions, such as compensatory action during [20] T. Hülsdünker, A. Mierau, C. Neeb, H. Kleinöder, and H. K. motion sickness, postural control failure from unilateral Strüder, “Cortical processes associated with continuous balance control as revealed by EEG spectral power,” Neurosci. Lett., 2015. vestibular loss (UVL) [42] or cerebellar stroke [43]. In these [21] G. Mochizuki, S. Boe, A. Marlin, and W. E. McIlroy, “Perturbation- complex postural conditions, incident adaptive and habituative evoked cortical activity reflects both the context and consequence of processes are activated within the CNS to ensure the postural instability,” Neuroscience, 2010. [22] A. L. Adkin, A. D. Campbell, R. Chua, and M. G. Carpenter, “The maintenance of upright posture and normative gait. Adding influence of postural threat on the cortical response to unpredictable HD-EEG to dynamic posturography measurement enables the and predictable postural perturbations,” Neurosci. Lett., 2008. commensurate measurement of CNS activity and dynamic [23] A. R. Sipp, J. T. Gwin, S. Makeig, and D. P. Ferris, “Loss of balance during balance beam walking elicits a multifocal theta band postural kinematics during adaptation and habituation to key electrocortical response,” J. Neurophysiol., 2013. postural control tasks. This invokes a deeper emphasis on the [24] K. Fujiwara, M. Maekawa, N. Kiyota, and C. Yaguchi, “Adaptation importance of further investigation into the adaptive and changes in dynamic postural control and contingent negative variation during backward disturbance by transient floor translation habitual processes implicated in CNS response to disease (i.e. in the elderly.,” J. Physiol. Anthropol., 2012. UVL, cerebral and/or cerebellar diseases), which would provide [25] T. Hülsdünker, A. Mierau, and H. K. Strüder, “Higher balance task key insight towards the identification of compensatory targets demands are associated with an increase in individual alpha peak for clinical intervention. frequency,” Front. Hum. Neurosci., 2016. [26] R. Friðriksdóttir et al., “Brain Processing During Postural Control – A Study Case,” in IFMBE Proceedings, 2020, vol. 76, pp. 1147– REFERENCES [27] S. Carlsö ö , How man moves: kinesiological studies and methods,. Heinemann, 1972. [1] F. B. Horak, “Postural orientation and equilibrium: What do we [28] C. C. Harro, A. Marquis, N. Piper, and C. Burdis, “Reliability and need to know about neural control of balance to prevent falls?,” in Validity of Force Platform Measures of Balance Impairment in Age and Ageing, 2006. Individuals With Parkinson Disease,” Phys. Ther., 2016. [2] A. Shumway-Cook and F. B. Horak, “Assessing the influence of [29] P. Schubert, M. Kirchner, D. Schmidtbleicher, and C. T. Haas, sensory interaction on balance. Suggestion from the field,” Phys. “About the structure of posturography: Sampling duration, Ther., 1986. parametrization, focus of attention (part I),” J. Biomed. Sci. Eng., [3] P.-A. Fransson, “Analysis of adaptation in human postural control,” Lund University, 2005. [30] D. A. Winter, Biomechanics and Motor Control of Human [4] J. M. Winters and P. E. Crago, Eds., Biomechanics and Neural Movement: Fourth Edition. 2009. Control of Posture and Movement, 1st ed. Springer-Verlag New [31] T. Yamamoto et al., “Universal and individual characteristics of York, 2000. postural sway during quiet standing in healthy young adults,” [5] H. C. Diener, J. Dichgans, F. Bootz, and M. Bacher, “Early Physiol. Rep., 2015. 8 [32] L. Chiari, L. Rocchi, and A. Cappello, “Stabilometric parameters are affected by anthropometry and foot placement,” Clin. Biomech., [33] P. Davidovits, “Static Forces,” in Physics in Biology and Medicine, Elsevier, 2019, pp. 1–20. [34] A. Delorme and S. Makeig, “EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis,” J. Neurosci. Methods, 2004. [35] J. Zhou, L. Lipsitz, D. Habtemariam, and B. Manor, “Sub-sensory vibratory noise augments the physiologic complexity of postural control in older adults,” J. Neuroeng. Rehabil., vol. 13, no. 1, p. 44, Dec. 2016. [36] A. Mierau, W. Klimesch, and J. Lefebvre, “State-dependent alpha peak frequency shifts: Experimental evidence, potential mechanisms and functional implications,” Neuroscience. 2017. [37] T. Solis-Escalante, J. van der Cruijsen, D. de Kam, J. van Kordelaar, V. Weerdesteyn, and A. C. Schouten, “Cortical dynamics during preparation and execution of reactive balance responses with distinct postural demands,” Neuroimage, vol. 188, pp. 557–571, Mar. 2019. [38] S. Slobounov, C. Cao, N. Jaiswal, and K. M. Newell, “Neural basis of postural instability identified by VTC and EEG,” Exp. Brain Res., [39] L. N. Saftari and O. S. Kwon, “Ageing vision and falls: A review,” Journal of Physiological Anthropology, vol. 37, no. 1. BioMed Central Ltd., 21-Feb-2018. [40] C. Zetterlund, L. O. Lundqvist, and H. O. Richter, “Visual, musculoskeletal and balance symptoms in individuals with visual impairment,” Clin. Exp. Optom., vol. 102, no. 1, pp. 63–69, Jan. [41] R. Johansson, P. A. Fransson, and M. Magnusson, “Optimal coordination and control of posture and movements,” J. Physiol. Paris, 2009. [42] R. J. Peterka, K. D. Statler, D. M. Wrisley, and F. B. Horak, “Postural compensation for unilateral vestibular loss,” Front. Neurol., 2011. [43] B. Manor et al., “Altered control of postural sway following cerebral infarction: A cross-sectional analysis,” Neurology, 2010. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Quantitative Biology arXiv (Cornell University)

Postural control adaptation and habituation during vibratory proprioceptive stimulation: an HD-EEG investigation of cortical recruitment and kinematics

Loading next page...
 
/lp/arxiv-cornell-university/postural-control-adaptation-and-habituation-during-vibratory-FuFO0e3yYY
ISSN
1534-4320
eISSN
ARCH-3345
DOI
10.1109/TNSRE.2020.2988585
Publisher site
See Article on Publisher Site

Abstract

Postural control adaptation and habituation during vibratory proprioceptive stimulation: an HD-EEG investigation of cortical recruitment and kinematics Fabio Barollo, Rún Friðriksdóttir, Kyle J. Edmunds, Gunnar H. Karlsson, Halldór Á. Svansson, Mahmoud Hassan, Antonio Fratini, Hannes Petersen and Paolo Gargiulo equilibrium condition of upright stance [1], [2]. Collectively Abstract— The objective of the present work is to measure defined as ‘postural control’ [3], this process is dynamically postural kinematics and power spectral variation from HD-EEG mediated by regulatory feedback elicited from somatosensory, to assess changes in cortical activity during adaptation and vestibular, and visual systems [4]. Exogenous disruption or habituation to postural perturbation. To evoke proprioceptive nonspecific stimulation of these systems can induce postural postural perturbation, vibratory stimulation at 85 Hz was applied sway [5], [6] – the magnitude and latency of which characterize to the calf muscles of 33 subjects over four 75-second stimulation the kinematics of postural control, which can be assessed by periods. Stimulation was performed according to a pseudorandom binary sequence. Vibratory impulses were synchronized to high- changes in force and torque actuated at the support surface of density electroencephalography (HD-EEG, 256 channels). the feet, altogether known as posturography [7], [8], [9]. Changes in absolute spectral power (ASP) were analyzed over four Postural control measurement is often employed under frequency bands (Δ: 0.5-3.5 Hz; θ: 3.5-7.5 Hz; α: 7.5-12.5 Hz; β: conditional balance perturbation, which is typically achieved 12.5-30 Hz). A force platform recorded torque actuated by the feet, via visual disturbance or proprioceptive stimulation [10]. and normalized sway path length (SPL) was computed as a Extant research in this regard cites postural control as a construct for postural performance during each period. SPL values indicated improvement in postural performance over the fundamental ‘learned’ motor skill, whose function and trial periods. Significant variation in absolute power values (ASP) efficiency can be systematically improved with routine postural was found in assessing postural adaptation: an increase in θ band tasks [9] or directed training [11]. From these studies, two ASP in the frontal-central region for closed-eyes trials, an increase dimensions of postural learning have been posited: in θ and β band ASP in the parietal region for open-eyes trials. In ‘adaptation’, defined as transient improvements in motor habituation, no significant variations in ASP were observed during response to upright balance perturbation [12], and ‘habituation’, closed-eyes trials, whereas an increase in θ, α, and β band ASP was observed with open eyes. Furthermore, open-eyed trials generally conversely defined by a gradual decrease in response to yielded a greater number of significant ASP differences across all repeated perturbation [13]. bands during both adaptation and habituation, suggesting that following cortical activity during postural perturbation may be While the response of sensorimotor systems during postural up-regulated with the availability of visual feedback. These results adaptation and habituation has been well-established in altogether provide deeper insight into pathological postural literature [14], [15] interrogating the commensurate role of the control failure by exploring the dynamic changes in both cortical activity and postural kinematics during adaptation and cerebral cortex or subcortical central nervous system (CNS) habituation to proprioceptive postural perturbation. structures is a comparatively recent subject of research. In this regard, literature has extended previous knowledge on Index Terms— Balance, cerebral cortex, HD-EEG, kinematics, subcortical balance maintenance [16], [17] to consider the postural control, power spectral density potential governing role of supratentorial information processing in the cerebral cortex [18]–[20]. The neuroimaging I. INTRODUCTION method of electroencephalography (EEG) has been cited for its UMAN posture is a complex and naturally unstable high temporal resolution in measuring cortical activity [6], [21], physiological process that requires the continuous [22]. In postural control research, balance perturbation has integration of compensatory mechanisms to maintain an Manuscript received MM DD, YY. Date of publication MM DD, YY. Fabio Barollo is with Reykjavik University and the Department of This research was supported jointly by the Institute for Biomedical and Biomedical Engineering at Aston University, Birmingham, UK. Neural Engineering at Reykjavík University, the Department of Anatomy at Antonio Fratini is with the Department of Biomedical Engineering at Aston the University of Iceland, and the Icelandic National Hospital (Landspítali University, Birmingham, UK. Scientific Fund) with additional funding support from the Rannís Icelandic M. Hassan is with LTSI- Université de Rennes, Rennes, France. Research Fund. Asterisk indicates corresponding author. H. Petersen is with The Department of Anatomy, University of Iceland, R. Friðriksdóttir, K. J. Edmunds, G.H. Karlsson, H.Á. Svansson and *P. Reykjavík, Iceland and with Akureyri Hospital, 600 Akureyri, Iceland Gargiulo are with Reykjavík University, Reykjavík, Iceland (Fabio Barollo and Rún Friðriksdóttir are co-first authors.) (correspondence e-mail: paolo@ru.is). 2 Fig. 1. Experimental set-up. revealed scalp-level activity changes in frontal-central and frontal-parietal cortical regions, specifically within α (7.5–12.5 II. MATERIALS AND METHODS Hz) and θ (3.5–7.5 Hz) frequency ranges [20], [23]. Additional A. Experimental setup EEG studies report bursts of γ activity (30–80 Hz) during voluntarily anterior-posterior movements [13]. EEG activity in As noted, the present experimental setup aimed at this regard is reported as changes in evoked time-domain event- integrating posturography measurement with the assessment of related potentials (ERP) or as perturbation-evoked responses cortical activity by 256-channel HD-EEG during vibratory (PERs), e.g. N1 amplitudes and Contingent Negative Variations proprioceptive stimulation. The study was reviewed and (CNV) [3], [6], [19], [24]. approved by the National Bioethics Committee (Vísindasiðanefnd - reference number: VSN-063) and the While recent evidence for the critical role of the cerebral measurements were performed at the Icelandic Center for cortex in governing postural adaptation and habituation has Neurophysiology at Reykjavik University. Thirty-three healthy been reported [10], to our knowledge, cortical activity volunteers (10 females and 23 males, aged 21 to 52) assessment during balance perturbation has never been participated in the study. These subjects had no history of synchronized with kinematic posturography measurements. In vertigo, central nervous disease, or lower extremity injury, and addition, no postural control studies report the use of ‘high- none of the subjects had consumed alcohol within a 24-hour density’, 256-channel EEG (HD-EEG) – a methodology with period prior to their measurement. Fig. 1 illustrates the overall superior spatial resolution to more conventional 32- or 64- experimental set-up for the present work, as presented at the XV channel systems. Furthermore, power spectral variation Mediterranean Conference on Medical and Biological analysis from EEG data remains underreported in postural Engineering and Computing – MEDICON 2019 [26]. control research, despite its being a conventional method for Participants were instructed to maintain an upright stance EEG signal analysis with proven utility in cognitive and motor during exogenous balance perturbation, evoked by the task studies [20], [25]. The present study aims to extend current simultaneous stimulation of vibrators fastened tightly by elastic research in this regard with the synchronized assessment of straps around the widest point of each calf. postural kinematics with power spectral variation analyses from The vibrators were designed using revolving DC-motors HD-EEG to quantify changes in cortical activity during equipped with a 3.5 gram eccentric weight, which was adaptation and habituation to postural perturbations using contained in a cylindrical casing approximately 6 cm in length vibratory proprioceptive stimulation. and 1 cm in diameter. Each stimulation was set to deliver 3 vibrations of 0.1 cm in amplitude, at a frequency of 85 Hz. III. DATA ANALYSIS Stimulation was applied according to a pseudorandom binary A. Assessing postural performance sequence schedule (PRBS), where each shift had a random duration of 1 to 6.6 seconds. Force platform data for both OE and CE datasets were B. Posturography measurement analogously segmented into five recording periods (QS, P1, P2, P3, and P4) in order to facilitate the synchronization of postural In general, maintaining a normative eased upright stance requires the symmetric distribution of body weight; when challenged, the resultant anterior-posterior or bilateral compensatory motion can be captured using a force platform to record changes in the body’s center of pressure [27], [28]. For the present work, this assessment was achieved using a customized platform system developed at the Department of Solid Mechanics, Lund Institute of Technology in Sweden [3]. Anterior-posterior (ant-post) and lateral (Lat) forces actuated by the feet were recorded at six degrees of freedom with an accuracy of 0.5 N; these data were sampled at 50 Hz by a custom-made program, Postcon™, on a computer equipped with an analogue-to-digital converter. Participants were instructed to stand on a pressure plate with their arms downwardly relaxed and their feet positioned at an angle of approximately 30 degrees, open to the front, with their heels approximately three centimeters apart. Participants were asked to focus on a fixed marker point in front of them, at about 150 Fig. 2. Statokinesigram showing an example subject’s body sway. Left: no cm distance. stimulation applied. Right: randomized vibratory stimulation applied on the calves to disrupt upright stance. Anterior- posterior and lateral displacements are expressed in cm. C. HD-EEG data acquisitions sway data with any evoked changes in cortical activity. The normalized Sway Path Length (SPL) was computed to describe HD-EEG data were acquired using a 256-channel Ag/AgCl the overall postural performance of the subjects during each wet-electrode cap connected in bipolar configuration to four period. cascaded 64-channel amplifiers; data collection employed a standardized 10-20 system montage with EEGO software (ANT The center of pressure (CoP) trajectory coincides with the neuro, Enschede Netherlands). An additional infra-orbital vertical projection on platform plane of the subject’s center of electrode was used to identify any obfuscating mass and it is widely used in human posture studies to assess electrooculographic (EoG) signals, and EEG data were body sway [29], [30]. Torque values were extracted from the continuously recorded at a sampling frequency of 1024 Hz. To collected force platform data and used to derive ant-post and lat synchronize EEG acquisition with posturography data, a displacement during each trial. A graphical representation of custom trigger signal box was built to rectify each vibratory the CoP trajectory (‘statokinesigram’) was then obtained by stimuli as a 5V TTL timing signal sent to the master amplifier. plotting the displacement along the ant-post and lat axes over This trigger system allowed for the generation of vibratory time, see Fig. 2. Sway Path Length is one of several on/off event timestamps at <1 ms latency during EEG posturographic parameters that can be extracted from a postural recordings. As our previous research has identified changes in platform and is one of the most commonly used [31]. cortical activity according to the availability of visual feedback, Nevertheless, previous research has established the sensitivity two measurement trials were performed for each subject, of these indices to different anthropometric characteristics – beginning with open eyes (OE) and followed by closed eyes particularly height and weight, where taller or heavier people (CE). An initial quiet stance (QS) baseline phase (30 seconds) appear to be more unstable [32]. As such, SPL values were preceded each stimulation phase (300 seconds), resulting in a normalized to each subject’s height and weight to estimate the total duration of 330 seconds for each recording. Each CoP trajectory on the platform according to the following stimulation phase was further subdivided into four 75-second formula: recording periods: P1, P2, P3, and P4. 𝑔𝑟𝑆𝑡𝑙𝑜𝑎𝑚𝑎𝑏𝑖 (𝑡 ) = ( ) ( ) 0.56∙ℎ ∙ 𝑚 ∙9.81 𝑘 𝑘 TABLE I Normalized Sway Path Length (SPL), means, and standard deviations by experimental epoch and OE/CE condition. QS P1 P2 P3 P4 Eyes Open Closed Open Closed Open Closed Open Closed Open Closed Mean 0.3038 0.4408 1.3624 1.8584 1.2190 1.7276 1.2312 1.7229 1.2574 1.6655 SD 0.0821 0.1392 0.2810 0.5130 0.2389 0.4447 0.2285 0.4204 0.2422 0.4285 Where 𝑘 refers to the individual participant, 0.56∙ℎ is an heteroscedastic t-tests, which yielded topological p-value maps estimate of the center of mass’s (CoM) radial distance to the for each EEG waveform (with p<0.05 the threshold for platform [33] and 𝑚 ∙9.81 is the weight applied to the significance). False discovery rate (FDR) and Bonferroni participant’s CoM. significance correction methods were employed and compared to address the statistical problem of multiple comparisons. B. HD-EEG data preprocessing Channels that resisted these multiple comparison correction methods were specifically marked in significance topologies. Raw HD-EEG data were pre-processed using the EEGLAB Toolbox [34], first with a band-pass filter set between 0.5–80 IV. RESULTS Hz, followed by a notch filter (49.5–50.5Hz) to remove AC A. SPL data distribution to assess adaptation and habituation power line interference from each period. Three approaches to periods artifact rejection were performed: channel interpolation, automatic continuous artifact rejection, and principle When describing time-varying phenomena, a distinction component analysis (PCA). All channels were re-referenced to must be made between true adaptation resulting from control a common average. dynamic alterations and other time-dependent phenomena, such as fatigue or biomechanical alterations resultant from changes To investigate variations in cortical activity between in balance conditions [3]. Because of this, we aimed to use SPL periods, absolute spectral power (ASP) values were obtained results to observe whether postural performance decreased over using fast Fourier transformation (FFT) analysis at a resolution any of the experimental periods; with both OE and CE of 0.977 Hz with a 10% Hanning window. This analysis was conditions, minimum postural sway has been reported to occur performed for four frequency bands: Δ (0.5-3.5 Hz), θ (3.5-7.5 at around 150-200 seconds following incident stimuli [3], [10]. Hz), α (7.5-12.5 Hz) and β (12.5-30 Hz). This analysis Table I shows the means and standard deviations of the generated ASP values for each frequency band and period, normalized SPL values for the present cohort, across each of which were extracted and exported for statistical analyses using the experimental periods and conditions. Fig. 3 represents the a customized Matlab GUI (MathWorks, Inc., Natick, 158 SPL data distribution with box plots to highlight median, mean Massachusetts, USA). From this analysis, topological maps and outlier values for each period. From these results, SPL did illustrating OE and CE difference spectra were extracted for increase in Period 4 (P4) for OE trials, but remained each EEG frequency band. In addition, significant changes in comparatively stable over all four periods for CE trials. To mean ASP at each electrode were assessed using paired Fig. 3.Normalized Sway Path Length. Left: Open Eyes; Right: Closed Eyes. avoid any potential obfuscation from fatigue or other time- Adaptation: dependent variations, we therefore used the difference between Period 3 (P3) and Period 1 (P1) ASP values in our assessment As previously mentioned, postural adaptation is shown of habituation. using normalized ASP differences between P1 and QS, and Figs. 4 and 5 depict cortical maps of channels showing significant changes in CE and OE conditions, respectively. B. Absolute power spectra variation from HD-EEG These results show an increase in ASP for both conditions, with many more significant regions present from OE trials. Figs. 4–6 show the results from statistical analyses Generally, ASP in the θ band increased in adaptation – performed across every channel. Each of the different colored particularly in the frontal-central region (p<0.05, FDR regions defines a topological map of statistical significance corrected) during CE and the parietal region (p<0.05, FDR (p<0.05) corrected for multiple comparisons in accordance with corrected) during OE, where ten electrodes passed the the following conditions: uncorrected paired t-test (yellow), Bonferroni correction test. During OE trials, higher frequency FDR correction (orange), and Bonferroni correction (red). bands (α and β) show significant activity as well. In particular, Channels that remained significant following FDR correction Fig. 5 shows increased activity in the parietal-occipital region are highlighted with a black ‘X’, while electrodes that resisted in the β band. Bonferroni correction are indicated by a blue ‘*’. White regions are considered non-significant (p>0.05). In the second row, Habituation: topographical maps of changes in ASP are shown solely for areas that presented a statistically significant difference Postural habituation is shown using normalized ASP between the examined periods. differences between P3 and P1. Fig. 6 depicts the cortical mapping of grand mean changes in ASP solely in channels shown to have a significant difference between periods Fig. 4. Adaptation CE. Top row: statistical analysis results. Each colored region identifies topological areas that achieved statistical significance (p<0.05) given the following conditions: uncorrected paired t-test (yellow), FDR correction (orange), and Bonferroni correction (red). Individual electrodes that resisted FDR correction are indicated by a black ‘X’. Bottom row: Topographies highlighting the average changes in ASP (between P1 and QS periods) over the whole cohort. Fig. 5. Adaptation OE. Top row: statistical analysis results. Each colored region identifies topological areas that achieved statisti cal significance (p<0.05) given the following conditions: uncorrected paired t-test (yellow), FDR correction (orange), and Bonferroni correction (red). Individual electrodes that resisted FDR correction are indicated by a black ‘X’, while electrodes that resisted Bonferroni correction are indicated by a blue ‘*’. Bottom row: Topographies highlighting the average changes in ASP (between P1 and QS periods) over the whole cohort. mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz mV /Hz 6 Fig. 6. Habituation OE. Top row: statistical analysis results. Each colored region identifies topological areas that achieved statistical significance (p<0.05) given the following conditions: uncorrected paired t-test (yellow), FDR correction (orange), and Bonferroni correction (red). Individual electrodes that resisted FDR correction are indicated by a black ‘X’. Bottom row: Topographies highlighting the average changes in ASP (between P3 and P1 periods) over the whole cohort. (p<0.05). White areas are again considered non-significant. The post direction. In all of these comparisons, no meaningful overall results show an increase of ASP in P3 compared to P1 correlation was found, suggesting the future importance of for OE trials, whereas during the CE periods, no significant considering alternative metrics for postural performance, such changes occurred. OE results specifically indicate increased as approximate entropy or multiscale entropy, which has shown activity in the θ band (temporal region), α band (parietal promise in linking modifications in neural involvement to region), and β band (frontal region). responses to vibration [35]. V. DISCUSSION Significant ASP differences were found across the entirety of the cortex in α and θ bands, with the exception of the prefrontal Power spectral variation analysis from EEG data remains area which yielded minimal significance in the α band. Power underreported in postural control literature. The present study in the θ band has been shown to increase in adaptation – aimed to extend current research with the synchronized particularly in the frontal-central region during CE trials and the assessment of postural kinematics with power spectral variation parietal area during OE trials [20], [23], [36]. Here, our results analyses from HD-EEG to quantify changes in cortical activity are in accordance with the recent work by Solis-Escalante et al., during adaptation and habituation during a postural control task who showed significant midline ASP differences in both CE using vibratory proprioceptive stimulation. and OE conditions as evidenced by the differential modulations Postural sway was recorded to obtain normalized SPL of α and low-γ rhythms [37]. This altogether suggests that values over five experimental periods, which were then utilized central region ASP increases during high-demand postural to classify postural adaptation and postural habituation as correction, such as balance maintenance without allowing differences in ASP between specific recording periods. In this corrective foot placement, as performed in the present study. regard, adaptation was defined as the difference between P1 and QS periods, in accordance with extant postural perturbation Furthermore, it has been suggested that increase in θ activity literature [3], [10], [12]. However, normalized SPL was used to in the frontal-central regions is involved in error detection and define postural habituation, where P1 ASP values were processing of postural stability during balance control [22], subtracted from the period with the lowest mean SPL to avoid [38]. As such, the θ band ASP differences shown here may the influence of time-dependent variations such as fatigue. signify the planning of corrective steps and/or the analysis of falling consequences, as indicated by our previous work on There are two main limitations to note from this cortical functional dynamics during postural control. Relatedly, methodology. Firstly, while it remains possible that the onset of significant ASP differences in the α band may reflect an habituation or fatigue may differ subject-to-subject, previous inhibition of error detection within the cingulate cortex due to studies have shown that ant-post and lat torque variations habituation [10]. during OE and CE trials reach a consistent minimum around 150-200 seconds following incident stimuli, corresponding The present results indicate that OE trials reflect a greater with P3 in the present work [3], [10]. Nevertheless, further number of significant differences in ASP across all bands investigation into quantitative indicators for the measurement during both adaptation and habituation. This suggests that of cortical habituation is recommended. Another following both acute and prolonged proprioceptive methodological limitation arises from demonstrating the perturbation, cortical activity may be up-regulated with the relationship between ASP and SPL. In this regard, we availability of visual feedback. These results generally support investigated whether there was any linear correlation between our prior hypothesis that the visual recognition of instability ASP and SPL or other metrics derived from CoP trajectory, may play a critical role in governing cortical processes requisite such as the root-mean-square (RMS) of excursion in the ant- mV /Hz mV /Hz mV /Hz mV /Hz 7 stabilization of human posture after a sudden disturbance: influence for postural control [10]. This hypothesis resulted from of rate and amplitude of displacement,” Exp. Brain Res., 1984. previous work demonstrating that all-type visual impairment is [6] K. L. Goh, S. Morris, W. L. Lee, A. Ring, and T. Tan, “Postural and associated with an increased risk for injurious falling [39], [40]. cortical responses following visual occlusion in standing and sitting tasks,” Exp. Brain Res., 2017. However, in addition to the impact of vision, it is also important [7] L. A. Cohen, “Role of eye and neck proprioceptive mechanisms in to note that postural control may be driven by brain network body orientation and motor coordination,” J. Neurophysiol., 1961. interactions rather than isolated changes in cortical activity at [8] M. H. Hu and M. H. Woollacott, “Multisensory training of standing balance in older adults: I. Postural stability and one-leg stance specific regions. Our present results support the potential balance,” Journals Gerontol., 1994. importance of network dynamics, as significant differences in [9] T. S. Galvão, E. S. Magalhães Júnior, M. A. Orsini Neves, and A. ASP were concurrently measured across many different cortical de Sá Ferreira, “Lower-limb muscle strength, static and dynamic regions. In this context, we believe that further work postural stabilities, risk of falling and fear of falling in polio survivors and healthy subjects,” Physiotherapy Theory and investigating the reconfiguration of cortical networks during Practice, 2018. adaptation and habituation could reveal new insights about how [10] K. J. Edmunds et al., “Cortical recruitment and functional dynamics functionally coordinated brain activity may dictate postural in postural control adaptation and habituation during vibratory proprioceptive stimulation,” J. Neural Eng., 2019. control. Such a model could include source-space or [11] A. S. Pollock, B. R. Durward, P. J. Rowe, and J. P. Paul, “What is connectivity-based analyses, as we have previously illustrated balance?,” Clin. Rehabil., vol. 14, no. 4, pp. 402–406, 2000. using lower density EEG [10]. Furthermore, since the examined [12] T. D. J. Welch and L. H. Ting, “Mechanisms of motor adaptation in reactive balance control,” PLoS One, 2014. measures of postural performance (SPL and RMS) were not [13] J. C. Eccles, “Learning in the motor system,” Prog. Brain Res., vol. able to highlight a significant correlation between kinematics 64, pp. 3–18, Jan. 1986. and cortical recruitment, a deeper investigation based on [14] D. A. E. Bolton and J. E. Misiaszek, “Compensatory balance reactions during forward and backward walking on a treadmill,” different postural parameters would be a fruitful area for further Gait Posture, 2012. work. [15] C. F. Honeycutt and T. Richard Nichols, “Disruption of cutaneous feedback alters magnitude but not direction of muscle responses to postural perturbations in the decerebrate cat,” Exp. Brain Res., 2010. VI. CONCLUSIONS [16] R. Magnus, “Some results of studies in the physiology of posture,” Dynamic posturography is an established method for Lancet, 1926. evaluating postural control. By delivering controlled disruption [17] C. S. Sherrington, “Flexion‐reflex of the limb, crossed extension‐ reflex, and reflex stepping and standing,” J. Physiol., 1910. in the form of calf vibration, we can examine central nervous [18] A. Mierau, B. Pester, T. Hülsdünker, K. Schiecke, H. K. Strüder, system (CNS) processing by associating body inertia and and H. Witte, “Cortical Correlates of Human Balance Control,” changes in upright stance [41]. This notion is true for simple Brain Topogr., 2017. [19] J. V. Jacobs and F. B. Horak, “Cortical control of postural upright stance but does not apply to complex postural tasks or responses,” in Journal of Neural Transmission, 2007. pathological conditions, such as compensatory action during [20] T. Hülsdünker, A. Mierau, C. Neeb, H. Kleinöder, and H. K. motion sickness, postural control failure from unilateral Strüder, “Cortical processes associated with continuous balance control as revealed by EEG spectral power,” Neurosci. Lett., 2015. vestibular loss (UVL) [42] or cerebellar stroke [43]. In these [21] G. Mochizuki, S. Boe, A. Marlin, and W. E. McIlroy, “Perturbation- complex postural conditions, incident adaptive and habituative evoked cortical activity reflects both the context and consequence of processes are activated within the CNS to ensure the postural instability,” Neuroscience, 2010. [22] A. L. Adkin, A. D. Campbell, R. Chua, and M. G. Carpenter, “The maintenance of upright posture and normative gait. Adding influence of postural threat on the cortical response to unpredictable HD-EEG to dynamic posturography measurement enables the and predictable postural perturbations,” Neurosci. Lett., 2008. commensurate measurement of CNS activity and dynamic [23] A. R. Sipp, J. T. Gwin, S. Makeig, and D. P. Ferris, “Loss of balance during balance beam walking elicits a multifocal theta band postural kinematics during adaptation and habituation to key electrocortical response,” J. Neurophysiol., 2013. postural control tasks. This invokes a deeper emphasis on the [24] K. Fujiwara, M. Maekawa, N. Kiyota, and C. Yaguchi, “Adaptation importance of further investigation into the adaptive and changes in dynamic postural control and contingent negative variation during backward disturbance by transient floor translation habitual processes implicated in CNS response to disease (i.e. in the elderly.,” J. Physiol. Anthropol., 2012. UVL, cerebral and/or cerebellar diseases), which would provide [25] T. Hülsdünker, A. Mierau, and H. K. Strüder, “Higher balance task key insight towards the identification of compensatory targets demands are associated with an increase in individual alpha peak for clinical intervention. frequency,” Front. Hum. Neurosci., 2016. [26] R. Friðriksdóttir et al., “Brain Processing During Postural Control – A Study Case,” in IFMBE Proceedings, 2020, vol. 76, pp. 1147– REFERENCES [27] S. Carlsö ö , How man moves: kinesiological studies and methods,. Heinemann, 1972. [1] F. B. Horak, “Postural orientation and equilibrium: What do we [28] C. C. Harro, A. Marquis, N. Piper, and C. Burdis, “Reliability and need to know about neural control of balance to prevent falls?,” in Validity of Force Platform Measures of Balance Impairment in Age and Ageing, 2006. Individuals With Parkinson Disease,” Phys. Ther., 2016. [2] A. Shumway-Cook and F. B. Horak, “Assessing the influence of [29] P. Schubert, M. Kirchner, D. Schmidtbleicher, and C. T. Haas, sensory interaction on balance. Suggestion from the field,” Phys. “About the structure of posturography: Sampling duration, Ther., 1986. parametrization, focus of attention (part I),” J. Biomed. Sci. Eng., [3] P.-A. Fransson, “Analysis of adaptation in human postural control,” Lund University, 2005. [30] D. A. Winter, Biomechanics and Motor Control of Human [4] J. M. Winters and P. E. Crago, Eds., Biomechanics and Neural Movement: Fourth Edition. 2009. Control of Posture and Movement, 1st ed. Springer-Verlag New [31] T. Yamamoto et al., “Universal and individual characteristics of York, 2000. postural sway during quiet standing in healthy young adults,” [5] H. C. Diener, J. Dichgans, F. Bootz, and M. Bacher, “Early Physiol. Rep., 2015. 8 [32] L. Chiari, L. Rocchi, and A. Cappello, “Stabilometric parameters are affected by anthropometry and foot placement,” Clin. Biomech., [33] P. Davidovits, “Static Forces,” in Physics in Biology and Medicine, Elsevier, 2019, pp. 1–20. [34] A. Delorme and S. Makeig, “EEGLAB: An open source toolbox for analysis of single-trial EEG dynamics including independent component analysis,” J. Neurosci. Methods, 2004. [35] J. Zhou, L. Lipsitz, D. Habtemariam, and B. Manor, “Sub-sensory vibratory noise augments the physiologic complexity of postural control in older adults,” J. Neuroeng. Rehabil., vol. 13, no. 1, p. 44, Dec. 2016. [36] A. Mierau, W. Klimesch, and J. Lefebvre, “State-dependent alpha peak frequency shifts: Experimental evidence, potential mechanisms and functional implications,” Neuroscience. 2017. [37] T. Solis-Escalante, J. van der Cruijsen, D. de Kam, J. van Kordelaar, V. Weerdesteyn, and A. C. Schouten, “Cortical dynamics during preparation and execution of reactive balance responses with distinct postural demands,” Neuroimage, vol. 188, pp. 557–571, Mar. 2019. [38] S. Slobounov, C. Cao, N. Jaiswal, and K. M. Newell, “Neural basis of postural instability identified by VTC and EEG,” Exp. Brain Res., [39] L. N. Saftari and O. S. Kwon, “Ageing vision and falls: A review,” Journal of Physiological Anthropology, vol. 37, no. 1. BioMed Central Ltd., 21-Feb-2018. [40] C. Zetterlund, L. O. Lundqvist, and H. O. Richter, “Visual, musculoskeletal and balance symptoms in individuals with visual impairment,” Clin. Exp. Optom., vol. 102, no. 1, pp. 63–69, Jan. [41] R. Johansson, P. A. Fransson, and M. Magnusson, “Optimal coordination and control of posture and movements,” J. Physiol. Paris, 2009. [42] R. J. Peterka, K. D. Statler, D. M. Wrisley, and F. B. Horak, “Postural compensation for unilateral vestibular loss,” Front. Neurol., 2011. [43] B. Manor et al., “Altered control of postural sway following cerebral infarction: A cross-sectional analysis,” Neurology, 2010.

Journal

Quantitative BiologyarXiv (Cornell University)

Published: Apr 19, 2020

There are no references for this article.