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

Learn More →

Towards the Enhancement of Body Standing Balance Recovery by Means of a Wireless Audio-Biofeedback System

Towards the Enhancement of Body Standing Balance Recovery by Means of a Wireless... Human maintain their body balance by sensorimotor controls mainly based on information gathered from vision, proprioception and vestibular systems. When there is a lack of information, caused by pathologies, diseases or aging, the subject may fall. In this context, we deve loped a system to augment information gathering, providing the subject with warning audio-feedback signals related to his/her equilibrium. The system comprises an inertial measurement unit (IMU), a data processing unit, a headphone audio device and a software appli cation. The IMU is a low-weight, small-size wireless instrument that, body-back located between the L2 and L5 lumbar vertebrae, measures the subject’s trunk kinematics. The application drives the data processing unit to feeding the headphone with electric signals related to the kinematic measures. Consequently, the user is audio-alerted, via headphone, of his/her own equilibrium, hearing a pleasant sound when in a stable equilibrium, or an increasing bothering sound when in an i ncreasing unstable condition. Tests were conducted on a group of six older subjects (59y-61y, SD = 2.09y) and a group of four young subjects (21y-26y, SD = 2.88y) to underline difference in effectiveness of the system, if any, related to the age of the users. For each subject, standing balance tests were performed in normal or altered conditions, such as, open or closed eyes, and on a solid or foam surface The system was evaluated in terms of usability, reliability, and effectiveness in improving the subject’s balance i n all conditions. As a result, the system successfully helped the subjects in reducing the body swaying within 10.65%-65.90%, differences depending on subjects’ age and test conditions. Keywords: Bio-feedback, Postural stability, E-rehabilitation, IMU, Standing balance 1. Introduction falling can psychological produce fear of falling again and depres - sion, which can lead to social isolation [5]. A sedentary lifestyle and drug usage can induce a slowdown of reflexes even in young people The visual, the proprioception and the vestibular systems work [4] (statistics not yet available). to provide standing balance [1,2], so that when one or more of When the sense of balance is reduced because of system is - these systems are under sufferance (because of pathologies, dis - sue(s), subjects tend to compensate by means of other systems, as it eases, or other reasons) balance capabilities can be reduce, even can be when information from the vestibular system increases in severely, in effectiveness. importance if the visual system is limited or absent. Key enabling The amount of reduction differs because of number/amount of technology (KET) can provide the subject with information no more the systems involved, and can be age -related (because of senes - or insufficiently supplied by the vestib ular system [6–10]. Within cence). this frame, here we propose an audio-biofeedback-based technology Pathologies of the ear, brain, or sensory nerve can cause dizzi - (ABF-T) as a KET useful to provide subjects with audio ness, which affects 47% of men and 61% of women over 70 years signals coded on the basis of his/her measured standing balance. of age [3], causing the 25% of falls [4]. Fall-related injuries in older The ABF-T consists of wearable devices, or wearables, which are people are a major global health problem, and consequences of a belt-worn inertial measurement unit (IMU) plus a headphone, and desktop hardware, which comprises a data receiving station and a personal computer. In particular, the wearables are battery - Corresponding author. E-mail address: saggio@uniroma2.it, wmdtg@email.it (G. powered low-weight, low-size, operator-independent, and low-cost Saggio). © 2019. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ apparatuses. The desktop hardware, wireless comm unicating with the wearables, provides A/D and D/A conversions, and real -time process data feeding signals to the headphone. The overall system is validated with ten subjects, grouped in young and older people, performing four different condition tests, so to take into account of normal, and altered-visual (i.e. with closed eyes), and of altered-proprioceptive (i.e. on a foam surface) conditions. 2. State of the art Within biofeedback systems [11–14], ABF is progressively ac- quiring greater relevance [15,16]. For example, ABF has been suc - cessfully exploited for motor rehabilitation after a hemiparetic stroke [17], when a patient was asked to move the hand to virtu - ally displace an object represented on a computer screen. A cone - like surface changed in color and a sound turned into musical notes according to the right or wrong trajectory of the patient’s hand. As a result, patients with sensory-motor deficits better performed compensatory movements using that system. In a recent study, Parkinson’s patients enhanced their balance by using an ABF-based therapy [18]. According to Petrofsky [19], ABF was used for the therapy of five subjects with impaired walking capability and reduced muscle strength of the hi p adductor (Trendelenburg’s sign), caused by spinal cord injury. By means of EMG sensors, the activity of the gluteus medium muscle was measured, and the pa - Fig. 1. (a) The Movit unit, dressed by the subject, which integrates a 3D gyroscope, tients were provided with audio signals coded to inform of incor - a 3D accelerometer, a 3D compass, a barometer, and a Li–Po battery, (b) the receiv- rect posture. The AB F was implem ented successfully also by Dozza ing unit, and the software termed “Captiks Motion Studio SDK, CMS”. et al. [20], who performed tests on healthy subjects in standing position, with closed eyes and on foam on top of a force plate. The The Movit unit (which can be networked with up to 15 other subjects’ balance information was related to stabilogram di ffusion units if necessary), wireless sends data to a receiving device (the and center of pressure analysis, from which a signal was conceived receiver) connected to a personal computer. Specifically, we ac - and audio- provided in a way that the subjects’ postural stability quired data from the internal 3-axis digital-output accelerometer, increased. which has a programmable full -scale range of ±2 g, ±4 g, ±8 g and Our ABS system evaluates the swaying characteristics of ±16 g, and integrates a 16-bit A/D converter, enabling simultane- standing subjects on a solid surface and on a foam surface, in both ous sampling of accelerometers without requiring an external mul- conditions of open and closed eyes. Innovatively, we empirically tiplexer. Data rate can range within 4 Hz –1000 Hz. We operated the defined different regions of swaying and audio -coding algorithms unit in low-power mode, at an operating current of 140 μA with a related to those regions, other from previously reported [21,22]. 50 Hz updating rate. The Movit was housed in a belt and located at the subject’s spine level, bet ween the second and the fifth lumbar 3. Materials and methods vertebra (L2 and L5, Fig. 2). We used a subset of two (out of three) parameters of the accelerometer, as useful for our purposes, in particular the pitch angle ( x coordinate, known as AP: The ABF system is based upon three logical components: a sen- sor unit (consisting of an inertial measurement unit, IMU, device) Anterior/Posterior), and the roll angle ( y coordinate, known as M L: Medial/Lateral). that measures the subject’s movements, a processing unit (consist- ing of an electronic device and a personal computer) that codes the A personal computer (i5 processor by Intel, 4GB Ram), con - sensory information in real time (routines written in Max/MSP, a nected to the receiver component, converts data into coded sounds Max Software Tools for Media, by Cycling ’74, San Francisco, USA), furnished to the subject by means of headphone ( model K701 by and an output unit (consisting of a headphone) that provides the AK G) with large-band frequency response (10Hz -39.800 Hz) and subject an acoustic representation of the physiological parameters. high dynamics (105 dB SPL/V). The headphone is an “open -type”, which allows the subject hearing the voice of an assistant, if nec - essary. 3.1. Hardware 3.2. Software The sensor unit is a belt-worn low-size (4.77 × 4.12 × 1.76 cm), low-weight (20 g), scientifically validated [23] IM U, termed Movit The personal computer runs proprietary software (termed Cap - (Fig. 1(a), by Captiks Srl, Rome, Italy), which provides inertial data tiks Motion Studio SD K, CM S, by Captiks Srl) for motion analysis related to the movements of the subject. The Movit is operated by a (Fig. 1(b)), so that pitch (x) and roll (y) signals (Fig. 3) are A/D con- 40 MHz clock microcontroller (AT32UC3A4256 by Atmel, San verted and further normalized, within 0 –1 range. Output data are Jose, California, US) and includes a 3D gyroscope, a 3D accelerom - converted into proper sound signals by means of routines based on eter, a 3D compass (each sensors integrated in a MEM S MPU-9150 Max/MSP (which is a graphical dataflow program ming environ - by Invensense, San Jose, California, US), and a barometer (BM P - ments for audio, by Cycling ’74, San Francisco, USA) [24]. 180 by Bosch Sensortec, Reutlingen, Germany). The Atmel micro - controller wireless sends data via 802.15.4 protocol, and a built-in 3.3. Regions and ranges of swaying Li-Po battery provides up to 8-hour supply, thus allowing no inter- rupti ons during tests (of only few minutes in our case). We modelled the body of the subject as an inverted pendulum, so that pitch ( x) and roll (y) of the trunk motion shaped a 7 2D orthogonally projected point P(x,y) on the floor (the coordi- nates of P expressed in degrees). By means of Max/MSP routines we generated two-channel signals related to the coordinates of the point P(x,y), accordingly to audio coding as defined in the follow - ing ( Section 3.4). We consider the swaying as the swing back-and-forth or to- and- from, i.e. the pitch and roll angles, as a distance dist of P(x,y) with respect to the position of P(x ,y ) itself when the subject naturally o o stands during the calibration phase, dist = P(x,y)-P(x ,y ), where x = o o x(t), y = y(t), and so dist = dist(x,y,t). Accordingly, P(x,y,t) can result within six different regions (from “A” to “F”) that we empirically identified in the 2D floor -plane, as schematized in Fig. 4. For each test (later discussed in Section 3.6) and for each parte- cipant (described Section 3.5), we determined the range R, as the difference between maximum and minimum distance dist(x,y,t), and the related variance V, both in absence and in presence of AB F (R , R and V , V , respectively). Finally, P and P noABF AB F noAB F ABF R V express such variations as percentages, as follows: P = R − R /R · 100 (1) R [( noABF ABF ) noABF] P = V − V /V · 100 (2) V [( noABF ABF ) noABF] Accordingly, P represents normalized differences between val - ues of the Range obtained without and with the ABF-T, respectively, so that the higher P the more effectiveness of the system. The regions of swaying identify “safety” (region “A”), “low-level warning” (region “B”), “medium-level warning” (region “C”), and “high-level warning” (regions “D”, “E”, “F”) conditions, and the audio-sounds are generated accordingly, in order to “reassuring” or “softly/hardly alerting” the subject. These regions are delimited by Fig. 2. The Movit was located at the subject’s spine level, between L2 and L5. Here, contours described by mathematical functions of circle, ellipse, and the subject stands on a rubber foam with closed eyes. rectangle, as reported in Table 1. Fig. 3. Pitch vs. time (a) and roll vs. time (b) for subject 2, with closed eyes, on foam rubber and without ABF. generator, which informs the subject that everything is cor- rectly working without alerts. The reference volume of the pink noise is set by the user him/her-self, according to preference. When the point P(x,y,t) is within Region “B”, a “low-level warn- ing” alerts the subject by an audio noise empirically filtered within 128Hz-14,263 Hz range, with a volume 1.5 higher than the reference. When P(x,y,t) is within Region “C”, an alerting noise is gener- ated, band- pass filtered within 415Hz-4390 Hz, with a volume three times higher than the reference. When P(x,y,t) is within Region “D”, an 800 Hz narrow-band fil- tered noise is generated. The frequency band is within a lower cut-off frequency f empirically set as an exponential function inf of y, f = 2^((8 + (y + 20)/40) 4, and a higher cut-off frequency inf sets as f = f + 800 Hz. In such a way, we provide the sub - su p inf jects with a pitch sound variation proportional to the tilt angle. With such a narrow band-pass filter, the subject perceives the sound as an almost pure sinusoidal wave. Based on the loga - rithmic human perception of pitch, we designed this perception to become stronger with frequency increasing. When P(x,y,t) is within Region “E” and “F”, we used the same approach as for Region “D”, i.e. 800 Hz narrow -band filtered noise, but modulated by a square wave panned to the headphone channels (left, x ≤ 2, for Region “E” and right, x ≥ 2, Fig. 4. Schematization of the six 2D different regions where the pitch (x) and roll for Region “F”), in order to obtain an intermittent noise. The (y) signals, coming from the IMU, are projected as a point P. Regions are divided among square wave ranged from 0 to 1 at a duty cycle of 50%, with the A: safety, B: low-level warning, C: medium-level warning, D, E, F: high-level warning, different in signal-balance between hears. period T exponentially decreasing in function of | x|, namely T = 8 , with exponent h = 2.5, h = 2, so that x x | |=0 | |=20 Table 1 T = 82 ms, T = 82.5 ms (h goes linearly from 2.5 to 2 when min ma x Identification of the six regions of swaying, from “A” to “F”, by |x| changes from 0 to 20). The resulting train of pulses had a means of their mathematical functions and levels of warning. frequency of about 15.6 Hz when |x| = 20, which also corre - Region Contour Warning? sponds to the maximum considered angle. 2 2 “A” x + y = 1 No (safety) 2 2 “B” [(y − 0.5)/2.25] + (x/1.5) = 1 Low-level Fig. 6 shows region-related cut-off frequencies. All signals were 2 2 “C” [(y − 0.5)/3] + (x/2) = 1 Medium-level wireless provided to the subjects so avoiding cables which could “D” − 2 < x < 2 and outside Region 3 High-level obstruct the body swaying. “E” x ≤ −2 High-level “F” x ≥ 2 High-level 3.5. Participants In order to compare balancing abilities between groups [25], Regions “B” and “C” are neither circular nor centered, since frontal tolerance of swaying is greater than backward or lateral tol - tests were conducted on 10 healthy subjects, divided into groups of young and older people. Specifically, four young subjects (21y-26y, erance of swaying in body balance (because of foot shape). SD = 2.88y) and six elder subjects (59y-61y, SD = 2.09y), equally di- vided into male and female. The study was approved by the local 3.4. Audio coding ethics com mittee, and all participants gave their informed consent in writing. Audio is coded from pink noise to an almost pure sine wave (Fig. 5 shows the intermediate occurrences), according to the severity of the unbalance status, and the sounds between the tw o 3.6. Test protocol hear channels are stereophonically bala nced or unbalanced in am - plitude. According to a previously validated protocol [8], the balance of In particular, according to the pitch values, when x < 0 (x > 0) each subject was considered in four different test conditions while the amplitude of the left channel increases at a given gain g(x), standing upright: (1) on floor with opened eyes; (2) on floor with while the amplitude of the right channel decreases, and vice versa, closed eyes; (3) on rubber foam layer with opened eyes; (4) on by the same amount. Audio coding was different for each region, rubber foam layer with closed eyes. Rubber foam layer and closed as detailed in the following: eyes conditions were considered to test deprived sensory conditions so simulating impairment at the proprioception level [26]. In When the point P(x,y,t) is within Region “A”, the subject is con- particular, the rubber foam is a 6-cm heightened AIRE X Balance - sidered to have a correct posture, with no risk of losing his/her pad plus foam carpet (FC; Airex AG, Sins, Switzerland) 25% balance. In such a condition, we considered that headphone’s compression resistance, 20 kPa; apparent density 55 kg m3; tensile silence could induce the perception of an unrealistic and annoy - strength 260 kPa). ing “absence of assistance”, which could even produce ambigu - All tests were performed in a quiet room, at a room temper - ity between “no warning” and “no working of the system.” On ature of 25 °C, with the subject, in turn, quietly with head still, the other end, a pure sine tone was bothersome for many par- while maintaining the upright standing position for 60 s. Each sub- ticipants. Therefore, similarly to telephone company providers ject was requested to reduce his/her body natural oscillation to a that generate a “courtesy noise”, we implemented a pink noise minimum. The four conditions were tested both with and without ABF, so resulting 4 × 2 = 8 (#tests x #replies) sessions for each Fig. 5. (a) Waveform and amplitude spectrum of the sound generated when (a) x = 0; y = 3 (filtered pink noise), (b) x = 0; y = 5 (mid freq. almost pure sine), (c) x = 0; y = 20 (high freq. almost pure sine). subject, while the sampling acquisition data rate was 50 Hz, so re - 4. Results and discussion sulting a total of (50 Hz x 60s = ) 3000 samples for each session. Before performing, each subject was trained for some minutes Fig. 7 eloquently shows the way point P “disperses” in eight dif - to get confident with the protocol. ferent occurrences of tests related to one older subject, combining Fig. 6. Upper and lower cut-off frequencies vs. pitch, according to the regions. Table 2 P and P values resulted from four different test conditions while subjects were, in turn, standing upright. Conditions R V were: on floor with opened eyes; on floor with closed eyes; on rubber foam layer with opened eyes; on rubber foam layer with closed eyes.n. Eyes open or closed floor or foam Older group Younger group Overall PR [%] PV [%] PR [%] PV [%] PR [%] PV [%] 10.65 29.63 56.04 65.55 26.92 49.24 34.44 58.10 53.53 64.74 40.62 58.10 31.97 45.76 49.21 65.90 35.16 50.68 29.87 26.19 38.27 46.23 36.61 37.86 situations of open/closed eyes, floor/foam platform, with/without When standing on the foam, the ABF results more effective for AB F. the young group when combined to the visual stimulus, the latter The ABF always increased the body standing balance, so that the being inborn with respect the few-minutes trained ABF (P R[%] = differences of (R R ) and (V − V ) always result in a noABF ABF noABF ABF 49.21 for open-eyes and P [%] = 38.27 for closed-eyes). positive value. P (Eq. (1)) and P (Eq. (2)) result positive, R V 5. Conclusions consequently. The increase in body balance, measured as the me- dian values of P and P , were computed for the elders (by means R V We designed an AB F-T based system and tested it, both on a of home-made routines written in Matlab R2017b, by Mathworks, group of six older people and on a group of four young subjects. Massachusetts, US), for the youngers and for the ensamble of the The system was able to measure the body swaying and to generate two groups, as schematized in Table 2. a sound corresponding to the measurements. This was to alert the According to our results, the increase in body balance is greater subject that his/her own balance was unstable with the potential for young than older subjects, but meaningful in both cases. Specif - consequence of falling. ically, the presence of the ABF improved the performance under The results demonstrated the effectiveness of AB F in helping conditions of closed eyes and standing on floor for the older group (P the subject in maintaining body balance, under different test con - = 34.44, P = 58.10), and conditions of opened eyes and standing on ditions. The distance of swaying in angle from the origin, namely floor for the young group (P = 56.04, P = 65.55). Differently, the R V the range R, and its variance V, were always lower when the A B F AB F in conditions of opened eyes and standing on floor was less was used with respect to their counterpart than when the ABF w as effective for the young group (P = 10.65, P = 29.63). Similar R V absent. These occurrences were confirmed in all the different test results are reported, among the same group, for all of the other conditions: four combinations of open/closed eyes and with/without conditions. foam rubber under the subject’s feet. When standing on the floor, the loss in visual feedback (closed- As a result, the ABF is capable to provide the subject with ad- eyes condition) makes the ABF useful in any case, but more relevant ditional information about his/her own somatosensory system to for older subjects than for you ng people (from P [%] = 10.62 to P [%] R R better control the body movements. We reported the worst results in = 34.44, and from P [%] = 56.04 to P [%] = 53.53, respectively). R R closed-eyes condition with foam rubber under the feet, as the total absence of visual and somatosensory information reduced Fig. 7. An exemplification of the “dispersion” of the point P on the base-plane related to the swaying assumed by one elderly subject when upright standing: (a) on floor, opened eyes, no ABF; (b) on floor, opened eyes, with ABF; (c) on floor, closed eyes, no ABF; (d) on floor, closed eyes, with ABF; (e) on foam, opened eyes, no ABF; (f) on foam, opened eyes, with ABF; (g) on foam, closed eyes, no ABF; (h) on foam, closed eyes, with ABF. the subject’s capacity to control his/her own movements. Both the open-eyes condition with foam rubber and the closed -eyes condi- Funding tion without foam rubber produced intermediate results. In con - The research was partially funded by the European Union’s ditions of closed eyes with the use of foam rub ber, the syste m Horizon 2020 research and innovation program me under the Marie proved to be more effective for the young subjects with respect to Sklodowska-Curie grant agreement No 659434. the older group of people. The ABF positively and differently augmented the balance re - Supplementary materials covery in every test condition and for every subject. Anyway, we argue that the higher effectiveness of the system on young people can be due to their higher reactiveness to external feedbacks, and Supplementary material associated with this article can be quicker response to multimedia stimuli (as, for instance, found, in the online version, at doi:10.1016/j.medengphy.2018.01. videogames can be). It can be also possible that a training period of few-minutes only, adopted to get confident with the system, played a rule too, since older people may claim a longer time to get used to References ABF. Finally, the ABF technology represents a low -weight, low-cost [1] Knoefel JE. Clinical neurology of aging. Oxford University Press; 2011. and non-intrusive system capable to enhance the subject’s stand- [2] Dewey D, Tupper DE. Developmental motor disorders: a neuropsychological ing balance. These characteristics suggest that ABT-T can augment perspective. The Guildford Press; 2004. traditional rehabilitation practice in a clinical setting or increase [3] Gazzola JM, Freitas Ganança F, Aratani MC, Rodrigues Perracini M, Malavasi Ganança M. Clinical evaluation of elderly people with chronic vestibular disorder. compliance in at-homebased postural exercise programs, success - Braz J Otorhinolaryngol 2006;72(4):515–22. fully. [4] WHO. Global report on falls prevention in older age. World Health Organiza- tion; 2008. [5] Lipsitz LA. The drop attack: a common geriatric symptom. J Am Geriatr 1983;31:617–20. [6] Costantini G, Carota M, Maccioni G, Giansanti D . A new integrated kinematic Ac kno wledgements sensor for the classification of sit-to-stand locomotion task. In: Proceedings of the IEEE international workshop on advances in sensors and interfaces; 2007. p. 38– Conflict of interest [7] Giansanti D, Morelli S, Maccioni G, Costantini G. Toward the design of a We declare no competing interests. wearable system for fall-risk detection in telerehabilitation. Telemed e-Health 2009;15(3):296–9. [8] Giansanti D, Maccioni G, Costantini G, Carota M. Towards the investigation of Ethics declarations kinematic parameters from an integrated measurement unit for the classification The study was approved by the local ethics com mittee of the rising from the chair. In: Proceedings of the IEEE EMBS annual international conference, New York City, USA; 2006. p. 1742–5. (Comi- tato Etico Indipendente, Tor Vergata, Protocollo 35/17). [9] Costantini G, Carota M, Maccioni G, Giansanti D. Discrimination between hu- [18] Mirelman A, Herman1 T, Nicolai S, Zijlstra A, Zijlstra W, Becker C, Chiari L, man functional ability/disability by means of different classification method- Hausdorff JM. Audio-Biofeedback training for posture and balance in Patients with ologies. In: Proceedings of the european conference on circuit theory and de- Parkinson’s disease. J NeuroEng Rehabil 2011;8(35):1–7. sign; 2007. p. 424–7. [19] Petrofsky JS. The use of electromyogram biofeedback to reduce Trendelenburg gait. [10] Saggio G, De Sanctis M, Cianca E, Latessa G, De Santis F, Giannini F. Long term Eur J Appl Physiol 2001;85(5):491–5. measurement of human joint movements for health care and rehabilitation [20] Dozza M, Chiari L, Chan B, Rocchi L, Horak F B, Cappello A. Influence of a purposes. wireless communications, vehicular technology, Proceedings of the portable audio-biofeedback device on structural properties of postural sway. J NeuroEng Rehabil 2005;2(13):1–12. information theory and aerospace & electronics systems Technology. Aalborg [21] Dozza M, Chiari L, Hlavacka F, Cappello A, Horak FB. Effects of linear versus (Denmark). 2009:674–678. sigmoid coding of visual or audio biofeedback for the control of upright stance. IEEE [11] James RE, Abarbanel A. Introduction to quantitative EEG and neurofeedback. Academic Press; 1999. Trans Neural Syst Rehabil Eng 2006;14(4):505–12. [12] Evans JR. The handbook of neurofeedback, dynamics and clinical applications. [22] Dozza M, Chiari L, Peterka RJ, Wall C, Horak FB. What is the most effective type of The Haworth Medical Press; 2007. audio-biofeedback for postural motor learning ? Gait Posture 2011;34(3):313–19. [13] Congedo M. Tomographic neurofeedback: a new technique for the self- [23] Alessandrini M, Micarelli A, Viziano A, Ottaviani F, Costantini G, Casali D, regulation of brain electrical activity. Dissertation Presented for the Philosophy Paolizzo F, Saggio G. Body-worn triaxial accelerometer coherence and reliabil- Doctor Degree, The University of Tennessee. 2003:17–18. ity related to static posturography in unilateral vestibular failure. Acta Otorhi- [14] Horak FB, Dozza M, Peterka R, Chiari L, Wall C. Vibrotactile biofeedback im - nolaryngol Italica 2017;37:231–6. proves tandem gait in patients with unilateral vestibular loss. Ann NY Acad Sci [24] Cycling74 Max/MSP, Documentation available on the web at: http://cycling74. 2009;1164:279–81. com/ [accessed 09.01.2018]. [15] Ghez C, Rikakis T, DuBois RL, Cook P. An auditory display system for aiding [25] Kiss RM. A new parameter for characterizing balancing ability on an unstable interjoint coordination. In: Proceedings of the international conference on auditory oscillatory platform. Med Eng Phys 2011;33(9):1160–6. display; 2000. [26] Rocchi L, Benocci M, Farella E, Benini L, Chiari L. Validation of a wireless [16] Costantini G, Saggio G, Casali D, Todisco M, Giansanti D, Maccioni G. Towards portable biofeedback system for balance control: preliminary results. In: Pro- ceedings of the second international conference on pervasive computing tech- the improvement of postural stability through audio biofeedback. In: IEEE in- nologies for healthcare; 2008. p. 254–7. ternational symposium on intelligent signal processing; 2015. p. 171–6. [17] Huang H, Ingalls T, Olson L, Ganley K, Rikakis T, He J. Interactive multimodal biofeedback for task-oriented neural rehabilitation. In: Proceedings of the in - ternational conference of IEEE engineering in medicine and biology society; 2005. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Electrical Engineering and Systems Science arXiv (Cornell University)

Towards the Enhancement of Body Standing Balance Recovery by Means of a Wireless Audio-Biofeedback System

Download PDF
8 pages

Loading next page...
 
/lp/arxiv-cornell-university/towards-the-enhancement-of-body-standing-balance-recovery-by-means-of-HHriY4FjVb

References (27)

ISSN
1350-4533
eISSN
ARCH-3348
DOI
10.1016/j.medengphy.2018.01.008
Publisher site
See Article on Publisher Site

Abstract

Human maintain their body balance by sensorimotor controls mainly based on information gathered from vision, proprioception and vestibular systems. When there is a lack of information, caused by pathologies, diseases or aging, the subject may fall. In this context, we deve loped a system to augment information gathering, providing the subject with warning audio-feedback signals related to his/her equilibrium. The system comprises an inertial measurement unit (IMU), a data processing unit, a headphone audio device and a software appli cation. The IMU is a low-weight, small-size wireless instrument that, body-back located between the L2 and L5 lumbar vertebrae, measures the subject’s trunk kinematics. The application drives the data processing unit to feeding the headphone with electric signals related to the kinematic measures. Consequently, the user is audio-alerted, via headphone, of his/her own equilibrium, hearing a pleasant sound when in a stable equilibrium, or an increasing bothering sound when in an i ncreasing unstable condition. Tests were conducted on a group of six older subjects (59y-61y, SD = 2.09y) and a group of four young subjects (21y-26y, SD = 2.88y) to underline difference in effectiveness of the system, if any, related to the age of the users. For each subject, standing balance tests were performed in normal or altered conditions, such as, open or closed eyes, and on a solid or foam surface The system was evaluated in terms of usability, reliability, and effectiveness in improving the subject’s balance i n all conditions. As a result, the system successfully helped the subjects in reducing the body swaying within 10.65%-65.90%, differences depending on subjects’ age and test conditions. Keywords: Bio-feedback, Postural stability, E-rehabilitation, IMU, Standing balance 1. Introduction falling can psychological produce fear of falling again and depres - sion, which can lead to social isolation [5]. A sedentary lifestyle and drug usage can induce a slowdown of reflexes even in young people The visual, the proprioception and the vestibular systems work [4] (statistics not yet available). to provide standing balance [1,2], so that when one or more of When the sense of balance is reduced because of system is - these systems are under sufferance (because of pathologies, dis - sue(s), subjects tend to compensate by means of other systems, as it eases, or other reasons) balance capabilities can be reduce, even can be when information from the vestibular system increases in severely, in effectiveness. importance if the visual system is limited or absent. Key enabling The amount of reduction differs because of number/amount of technology (KET) can provide the subject with information no more the systems involved, and can be age -related (because of senes - or insufficiently supplied by the vestib ular system [6–10]. Within cence). this frame, here we propose an audio-biofeedback-based technology Pathologies of the ear, brain, or sensory nerve can cause dizzi - (ABF-T) as a KET useful to provide subjects with audio ness, which affects 47% of men and 61% of women over 70 years signals coded on the basis of his/her measured standing balance. of age [3], causing the 25% of falls [4]. Fall-related injuries in older The ABF-T consists of wearable devices, or wearables, which are people are a major global health problem, and consequences of a belt-worn inertial measurement unit (IMU) plus a headphone, and desktop hardware, which comprises a data receiving station and a personal computer. In particular, the wearables are battery - Corresponding author. E-mail address: saggio@uniroma2.it, wmdtg@email.it (G. powered low-weight, low-size, operator-independent, and low-cost Saggio). © 2019. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ apparatuses. The desktop hardware, wireless comm unicating with the wearables, provides A/D and D/A conversions, and real -time process data feeding signals to the headphone. The overall system is validated with ten subjects, grouped in young and older people, performing four different condition tests, so to take into account of normal, and altered-visual (i.e. with closed eyes), and of altered-proprioceptive (i.e. on a foam surface) conditions. 2. State of the art Within biofeedback systems [11–14], ABF is progressively ac- quiring greater relevance [15,16]. For example, ABF has been suc - cessfully exploited for motor rehabilitation after a hemiparetic stroke [17], when a patient was asked to move the hand to virtu - ally displace an object represented on a computer screen. A cone - like surface changed in color and a sound turned into musical notes according to the right or wrong trajectory of the patient’s hand. As a result, patients with sensory-motor deficits better performed compensatory movements using that system. In a recent study, Parkinson’s patients enhanced their balance by using an ABF-based therapy [18]. According to Petrofsky [19], ABF was used for the therapy of five subjects with impaired walking capability and reduced muscle strength of the hi p adductor (Trendelenburg’s sign), caused by spinal cord injury. By means of EMG sensors, the activity of the gluteus medium muscle was measured, and the pa - Fig. 1. (a) The Movit unit, dressed by the subject, which integrates a 3D gyroscope, tients were provided with audio signals coded to inform of incor - a 3D accelerometer, a 3D compass, a barometer, and a Li–Po battery, (b) the receiv- rect posture. The AB F was implem ented successfully also by Dozza ing unit, and the software termed “Captiks Motion Studio SDK, CMS”. et al. [20], who performed tests on healthy subjects in standing position, with closed eyes and on foam on top of a force plate. The The Movit unit (which can be networked with up to 15 other subjects’ balance information was related to stabilogram di ffusion units if necessary), wireless sends data to a receiving device (the and center of pressure analysis, from which a signal was conceived receiver) connected to a personal computer. Specifically, we ac - and audio- provided in a way that the subjects’ postural stability quired data from the internal 3-axis digital-output accelerometer, increased. which has a programmable full -scale range of ±2 g, ±4 g, ±8 g and Our ABS system evaluates the swaying characteristics of ±16 g, and integrates a 16-bit A/D converter, enabling simultane- standing subjects on a solid surface and on a foam surface, in both ous sampling of accelerometers without requiring an external mul- conditions of open and closed eyes. Innovatively, we empirically tiplexer. Data rate can range within 4 Hz –1000 Hz. We operated the defined different regions of swaying and audio -coding algorithms unit in low-power mode, at an operating current of 140 μA with a related to those regions, other from previously reported [21,22]. 50 Hz updating rate. The Movit was housed in a belt and located at the subject’s spine level, bet ween the second and the fifth lumbar 3. Materials and methods vertebra (L2 and L5, Fig. 2). We used a subset of two (out of three) parameters of the accelerometer, as useful for our purposes, in particular the pitch angle ( x coordinate, known as AP: The ABF system is based upon three logical components: a sen- sor unit (consisting of an inertial measurement unit, IMU, device) Anterior/Posterior), and the roll angle ( y coordinate, known as M L: Medial/Lateral). that measures the subject’s movements, a processing unit (consist- ing of an electronic device and a personal computer) that codes the A personal computer (i5 processor by Intel, 4GB Ram), con - sensory information in real time (routines written in Max/MSP, a nected to the receiver component, converts data into coded sounds Max Software Tools for Media, by Cycling ’74, San Francisco, USA), furnished to the subject by means of headphone ( model K701 by and an output unit (consisting of a headphone) that provides the AK G) with large-band frequency response (10Hz -39.800 Hz) and subject an acoustic representation of the physiological parameters. high dynamics (105 dB SPL/V). The headphone is an “open -type”, which allows the subject hearing the voice of an assistant, if nec - essary. 3.1. Hardware 3.2. Software The sensor unit is a belt-worn low-size (4.77 × 4.12 × 1.76 cm), low-weight (20 g), scientifically validated [23] IM U, termed Movit The personal computer runs proprietary software (termed Cap - (Fig. 1(a), by Captiks Srl, Rome, Italy), which provides inertial data tiks Motion Studio SD K, CM S, by Captiks Srl) for motion analysis related to the movements of the subject. The Movit is operated by a (Fig. 1(b)), so that pitch (x) and roll (y) signals (Fig. 3) are A/D con- 40 MHz clock microcontroller (AT32UC3A4256 by Atmel, San verted and further normalized, within 0 –1 range. Output data are Jose, California, US) and includes a 3D gyroscope, a 3D accelerom - converted into proper sound signals by means of routines based on eter, a 3D compass (each sensors integrated in a MEM S MPU-9150 Max/MSP (which is a graphical dataflow program ming environ - by Invensense, San Jose, California, US), and a barometer (BM P - ments for audio, by Cycling ’74, San Francisco, USA) [24]. 180 by Bosch Sensortec, Reutlingen, Germany). The Atmel micro - controller wireless sends data via 802.15.4 protocol, and a built-in 3.3. Regions and ranges of swaying Li-Po battery provides up to 8-hour supply, thus allowing no inter- rupti ons during tests (of only few minutes in our case). We modelled the body of the subject as an inverted pendulum, so that pitch ( x) and roll (y) of the trunk motion shaped a 7 2D orthogonally projected point P(x,y) on the floor (the coordi- nates of P expressed in degrees). By means of Max/MSP routines we generated two-channel signals related to the coordinates of the point P(x,y), accordingly to audio coding as defined in the follow - ing ( Section 3.4). We consider the swaying as the swing back-and-forth or to- and- from, i.e. the pitch and roll angles, as a distance dist of P(x,y) with respect to the position of P(x ,y ) itself when the subject naturally o o stands during the calibration phase, dist = P(x,y)-P(x ,y ), where x = o o x(t), y = y(t), and so dist = dist(x,y,t). Accordingly, P(x,y,t) can result within six different regions (from “A” to “F”) that we empirically identified in the 2D floor -plane, as schematized in Fig. 4. For each test (later discussed in Section 3.6) and for each parte- cipant (described Section 3.5), we determined the range R, as the difference between maximum and minimum distance dist(x,y,t), and the related variance V, both in absence and in presence of AB F (R , R and V , V , respectively). Finally, P and P noABF AB F noAB F ABF R V express such variations as percentages, as follows: P = R − R /R · 100 (1) R [( noABF ABF ) noABF] P = V − V /V · 100 (2) V [( noABF ABF ) noABF] Accordingly, P represents normalized differences between val - ues of the Range obtained without and with the ABF-T, respectively, so that the higher P the more effectiveness of the system. The regions of swaying identify “safety” (region “A”), “low-level warning” (region “B”), “medium-level warning” (region “C”), and “high-level warning” (regions “D”, “E”, “F”) conditions, and the audio-sounds are generated accordingly, in order to “reassuring” or “softly/hardly alerting” the subject. These regions are delimited by Fig. 2. The Movit was located at the subject’s spine level, between L2 and L5. Here, contours described by mathematical functions of circle, ellipse, and the subject stands on a rubber foam with closed eyes. rectangle, as reported in Table 1. Fig. 3. Pitch vs. time (a) and roll vs. time (b) for subject 2, with closed eyes, on foam rubber and without ABF. generator, which informs the subject that everything is cor- rectly working without alerts. The reference volume of the pink noise is set by the user him/her-self, according to preference. When the point P(x,y,t) is within Region “B”, a “low-level warn- ing” alerts the subject by an audio noise empirically filtered within 128Hz-14,263 Hz range, with a volume 1.5 higher than the reference. When P(x,y,t) is within Region “C”, an alerting noise is gener- ated, band- pass filtered within 415Hz-4390 Hz, with a volume three times higher than the reference. When P(x,y,t) is within Region “D”, an 800 Hz narrow-band fil- tered noise is generated. The frequency band is within a lower cut-off frequency f empirically set as an exponential function inf of y, f = 2^((8 + (y + 20)/40) 4, and a higher cut-off frequency inf sets as f = f + 800 Hz. In such a way, we provide the sub - su p inf jects with a pitch sound variation proportional to the tilt angle. With such a narrow band-pass filter, the subject perceives the sound as an almost pure sinusoidal wave. Based on the loga - rithmic human perception of pitch, we designed this perception to become stronger with frequency increasing. When P(x,y,t) is within Region “E” and “F”, we used the same approach as for Region “D”, i.e. 800 Hz narrow -band filtered noise, but modulated by a square wave panned to the headphone channels (left, x ≤ 2, for Region “E” and right, x ≥ 2, Fig. 4. Schematization of the six 2D different regions where the pitch (x) and roll for Region “F”), in order to obtain an intermittent noise. The (y) signals, coming from the IMU, are projected as a point P. Regions are divided among square wave ranged from 0 to 1 at a duty cycle of 50%, with the A: safety, B: low-level warning, C: medium-level warning, D, E, F: high-level warning, different in signal-balance between hears. period T exponentially decreasing in function of | x|, namely T = 8 , with exponent h = 2.5, h = 2, so that x x | |=0 | |=20 Table 1 T = 82 ms, T = 82.5 ms (h goes linearly from 2.5 to 2 when min ma x Identification of the six regions of swaying, from “A” to “F”, by |x| changes from 0 to 20). The resulting train of pulses had a means of their mathematical functions and levels of warning. frequency of about 15.6 Hz when |x| = 20, which also corre - Region Contour Warning? sponds to the maximum considered angle. 2 2 “A” x + y = 1 No (safety) 2 2 “B” [(y − 0.5)/2.25] + (x/1.5) = 1 Low-level Fig. 6 shows region-related cut-off frequencies. All signals were 2 2 “C” [(y − 0.5)/3] + (x/2) = 1 Medium-level wireless provided to the subjects so avoiding cables which could “D” − 2 < x < 2 and outside Region 3 High-level obstruct the body swaying. “E” x ≤ −2 High-level “F” x ≥ 2 High-level 3.5. Participants In order to compare balancing abilities between groups [25], Regions “B” and “C” are neither circular nor centered, since frontal tolerance of swaying is greater than backward or lateral tol - tests were conducted on 10 healthy subjects, divided into groups of young and older people. Specifically, four young subjects (21y-26y, erance of swaying in body balance (because of foot shape). SD = 2.88y) and six elder subjects (59y-61y, SD = 2.09y), equally di- vided into male and female. The study was approved by the local 3.4. Audio coding ethics com mittee, and all participants gave their informed consent in writing. Audio is coded from pink noise to an almost pure sine wave (Fig. 5 shows the intermediate occurrences), according to the severity of the unbalance status, and the sounds between the tw o 3.6. Test protocol hear channels are stereophonically bala nced or unbalanced in am - plitude. According to a previously validated protocol [8], the balance of In particular, according to the pitch values, when x < 0 (x > 0) each subject was considered in four different test conditions while the amplitude of the left channel increases at a given gain g(x), standing upright: (1) on floor with opened eyes; (2) on floor with while the amplitude of the right channel decreases, and vice versa, closed eyes; (3) on rubber foam layer with opened eyes; (4) on by the same amount. Audio coding was different for each region, rubber foam layer with closed eyes. Rubber foam layer and closed as detailed in the following: eyes conditions were considered to test deprived sensory conditions so simulating impairment at the proprioception level [26]. In When the point P(x,y,t) is within Region “A”, the subject is con- particular, the rubber foam is a 6-cm heightened AIRE X Balance - sidered to have a correct posture, with no risk of losing his/her pad plus foam carpet (FC; Airex AG, Sins, Switzerland) 25% balance. In such a condition, we considered that headphone’s compression resistance, 20 kPa; apparent density 55 kg m3; tensile silence could induce the perception of an unrealistic and annoy - strength 260 kPa). ing “absence of assistance”, which could even produce ambigu - All tests were performed in a quiet room, at a room temper - ity between “no warning” and “no working of the system.” On ature of 25 °C, with the subject, in turn, quietly with head still, the other end, a pure sine tone was bothersome for many par- while maintaining the upright standing position for 60 s. Each sub- ticipants. Therefore, similarly to telephone company providers ject was requested to reduce his/her body natural oscillation to a that generate a “courtesy noise”, we implemented a pink noise minimum. The four conditions were tested both with and without ABF, so resulting 4 × 2 = 8 (#tests x #replies) sessions for each Fig. 5. (a) Waveform and amplitude spectrum of the sound generated when (a) x = 0; y = 3 (filtered pink noise), (b) x = 0; y = 5 (mid freq. almost pure sine), (c) x = 0; y = 20 (high freq. almost pure sine). subject, while the sampling acquisition data rate was 50 Hz, so re - 4. Results and discussion sulting a total of (50 Hz x 60s = ) 3000 samples for each session. Before performing, each subject was trained for some minutes Fig. 7 eloquently shows the way point P “disperses” in eight dif - to get confident with the protocol. ferent occurrences of tests related to one older subject, combining Fig. 6. Upper and lower cut-off frequencies vs. pitch, according to the regions. Table 2 P and P values resulted from four different test conditions while subjects were, in turn, standing upright. Conditions R V were: on floor with opened eyes; on floor with closed eyes; on rubber foam layer with opened eyes; on rubber foam layer with closed eyes.n. Eyes open or closed floor or foam Older group Younger group Overall PR [%] PV [%] PR [%] PV [%] PR [%] PV [%] 10.65 29.63 56.04 65.55 26.92 49.24 34.44 58.10 53.53 64.74 40.62 58.10 31.97 45.76 49.21 65.90 35.16 50.68 29.87 26.19 38.27 46.23 36.61 37.86 situations of open/closed eyes, floor/foam platform, with/without When standing on the foam, the ABF results more effective for AB F. the young group when combined to the visual stimulus, the latter The ABF always increased the body standing balance, so that the being inborn with respect the few-minutes trained ABF (P R[%] = differences of (R R ) and (V − V ) always result in a noABF ABF noABF ABF 49.21 for open-eyes and P [%] = 38.27 for closed-eyes). positive value. P (Eq. (1)) and P (Eq. (2)) result positive, R V 5. Conclusions consequently. The increase in body balance, measured as the me- dian values of P and P , were computed for the elders (by means R V We designed an AB F-T based system and tested it, both on a of home-made routines written in Matlab R2017b, by Mathworks, group of six older people and on a group of four young subjects. Massachusetts, US), for the youngers and for the ensamble of the The system was able to measure the body swaying and to generate two groups, as schematized in Table 2. a sound corresponding to the measurements. This was to alert the According to our results, the increase in body balance is greater subject that his/her own balance was unstable with the potential for young than older subjects, but meaningful in both cases. Specif - consequence of falling. ically, the presence of the ABF improved the performance under The results demonstrated the effectiveness of AB F in helping conditions of closed eyes and standing on floor for the older group (P the subject in maintaining body balance, under different test con - = 34.44, P = 58.10), and conditions of opened eyes and standing on ditions. The distance of swaying in angle from the origin, namely floor for the young group (P = 56.04, P = 65.55). Differently, the R V the range R, and its variance V, were always lower when the A B F AB F in conditions of opened eyes and standing on floor was less was used with respect to their counterpart than when the ABF w as effective for the young group (P = 10.65, P = 29.63). Similar R V absent. These occurrences were confirmed in all the different test results are reported, among the same group, for all of the other conditions: four combinations of open/closed eyes and with/without conditions. foam rubber under the subject’s feet. When standing on the floor, the loss in visual feedback (closed- As a result, the ABF is capable to provide the subject with ad- eyes condition) makes the ABF useful in any case, but more relevant ditional information about his/her own somatosensory system to for older subjects than for you ng people (from P [%] = 10.62 to P [%] R R better control the body movements. We reported the worst results in = 34.44, and from P [%] = 56.04 to P [%] = 53.53, respectively). R R closed-eyes condition with foam rubber under the feet, as the total absence of visual and somatosensory information reduced Fig. 7. An exemplification of the “dispersion” of the point P on the base-plane related to the swaying assumed by one elderly subject when upright standing: (a) on floor, opened eyes, no ABF; (b) on floor, opened eyes, with ABF; (c) on floor, closed eyes, no ABF; (d) on floor, closed eyes, with ABF; (e) on foam, opened eyes, no ABF; (f) on foam, opened eyes, with ABF; (g) on foam, closed eyes, no ABF; (h) on foam, closed eyes, with ABF. the subject’s capacity to control his/her own movements. Both the open-eyes condition with foam rubber and the closed -eyes condi- Funding tion without foam rubber produced intermediate results. In con - The research was partially funded by the European Union’s ditions of closed eyes with the use of foam rub ber, the syste m Horizon 2020 research and innovation program me under the Marie proved to be more effective for the young subjects with respect to Sklodowska-Curie grant agreement No 659434. the older group of people. The ABF positively and differently augmented the balance re - Supplementary materials covery in every test condition and for every subject. Anyway, we argue that the higher effectiveness of the system on young people can be due to their higher reactiveness to external feedbacks, and Supplementary material associated with this article can be quicker response to multimedia stimuli (as, for instance, found, in the online version, at doi:10.1016/j.medengphy.2018.01. videogames can be). It can be also possible that a training period of few-minutes only, adopted to get confident with the system, played a rule too, since older people may claim a longer time to get used to References ABF. Finally, the ABF technology represents a low -weight, low-cost [1] Knoefel JE. Clinical neurology of aging. Oxford University Press; 2011. and non-intrusive system capable to enhance the subject’s stand- [2] Dewey D, Tupper DE. Developmental motor disorders: a neuropsychological ing balance. These characteristics suggest that ABT-T can augment perspective. The Guildford Press; 2004. traditional rehabilitation practice in a clinical setting or increase [3] Gazzola JM, Freitas Ganança F, Aratani MC, Rodrigues Perracini M, Malavasi Ganança M. Clinical evaluation of elderly people with chronic vestibular disorder. compliance in at-homebased postural exercise programs, success - Braz J Otorhinolaryngol 2006;72(4):515–22. fully. [4] WHO. Global report on falls prevention in older age. World Health Organiza- tion; 2008. [5] Lipsitz LA. The drop attack: a common geriatric symptom. J Am Geriatr 1983;31:617–20. [6] Costantini G, Carota M, Maccioni G, Giansanti D . A new integrated kinematic Ac kno wledgements sensor for the classification of sit-to-stand locomotion task. In: Proceedings of the IEEE international workshop on advances in sensors and interfaces; 2007. p. 38– Conflict of interest [7] Giansanti D, Morelli S, Maccioni G, Costantini G. Toward the design of a We declare no competing interests. wearable system for fall-risk detection in telerehabilitation. Telemed e-Health 2009;15(3):296–9. [8] Giansanti D, Maccioni G, Costantini G, Carota M. Towards the investigation of Ethics declarations kinematic parameters from an integrated measurement unit for the classification The study was approved by the local ethics com mittee of the rising from the chair. In: Proceedings of the IEEE EMBS annual international conference, New York City, USA; 2006. p. 1742–5. (Comi- tato Etico Indipendente, Tor Vergata, Protocollo 35/17). [9] Costantini G, Carota M, Maccioni G, Giansanti D. Discrimination between hu- [18] Mirelman A, Herman1 T, Nicolai S, Zijlstra A, Zijlstra W, Becker C, Chiari L, man functional ability/disability by means of different classification method- Hausdorff JM. Audio-Biofeedback training for posture and balance in Patients with ologies. In: Proceedings of the european conference on circuit theory and de- Parkinson’s disease. J NeuroEng Rehabil 2011;8(35):1–7. sign; 2007. p. 424–7. [19] Petrofsky JS. The use of electromyogram biofeedback to reduce Trendelenburg gait. [10] Saggio G, De Sanctis M, Cianca E, Latessa G, De Santis F, Giannini F. Long term Eur J Appl Physiol 2001;85(5):491–5. measurement of human joint movements for health care and rehabilitation [20] Dozza M, Chiari L, Chan B, Rocchi L, Horak F B, Cappello A. Influence of a purposes. wireless communications, vehicular technology, Proceedings of the portable audio-biofeedback device on structural properties of postural sway. J NeuroEng Rehabil 2005;2(13):1–12. information theory and aerospace & electronics systems Technology. Aalborg [21] Dozza M, Chiari L, Hlavacka F, Cappello A, Horak FB. Effects of linear versus (Denmark). 2009:674–678. sigmoid coding of visual or audio biofeedback for the control of upright stance. IEEE [11] James RE, Abarbanel A. Introduction to quantitative EEG and neurofeedback. Academic Press; 1999. Trans Neural Syst Rehabil Eng 2006;14(4):505–12. [12] Evans JR. The handbook of neurofeedback, dynamics and clinical applications. [22] Dozza M, Chiari L, Peterka RJ, Wall C, Horak FB. What is the most effective type of The Haworth Medical Press; 2007. audio-biofeedback for postural motor learning ? Gait Posture 2011;34(3):313–19. [13] Congedo M. Tomographic neurofeedback: a new technique for the self- [23] Alessandrini M, Micarelli A, Viziano A, Ottaviani F, Costantini G, Casali D, regulation of brain electrical activity. Dissertation Presented for the Philosophy Paolizzo F, Saggio G. Body-worn triaxial accelerometer coherence and reliabil- Doctor Degree, The University of Tennessee. 2003:17–18. ity related to static posturography in unilateral vestibular failure. Acta Otorhi- [14] Horak FB, Dozza M, Peterka R, Chiari L, Wall C. Vibrotactile biofeedback im - nolaryngol Italica 2017;37:231–6. proves tandem gait in patients with unilateral vestibular loss. Ann NY Acad Sci [24] Cycling74 Max/MSP, Documentation available on the web at: http://cycling74. 2009;1164:279–81. com/ [accessed 09.01.2018]. [15] Ghez C, Rikakis T, DuBois RL, Cook P. An auditory display system for aiding [25] Kiss RM. A new parameter for characterizing balancing ability on an unstable interjoint coordination. In: Proceedings of the international conference on auditory oscillatory platform. Med Eng Phys 2011;33(9):1160–6. display; 2000. [26] Rocchi L, Benocci M, Farella E, Benini L, Chiari L. Validation of a wireless [16] Costantini G, Saggio G, Casali D, Todisco M, Giansanti D, Maccioni G. Towards portable biofeedback system for balance control: preliminary results. In: Pro- ceedings of the second international conference on pervasive computing tech- the improvement of postural stability through audio biofeedback. In: IEEE in- nologies for healthcare; 2008. p. 254–7. ternational symposium on intelligent signal processing; 2015. p. 171–6. [17] Huang H, Ingalls T, Olson L, Ganley K, Rikakis T, He J. Interactive multimodal biofeedback for task-oriented neural rehabilitation. In: Proceedings of the in - ternational conference of IEEE engineering in medicine and biology society; 2005.

Journal

Electrical Engineering and Systems Science – arXiv (Cornell University)

Published: Jul 26, 2019

There are no references for this article.