An Experimental Paradigm to Assess Postural Stabilization: No More Movement and Not Yet Posture

Share Embed

Descrição do Produto



An Experimental Paradigm to Assess Postural Stabilization: No More Movement and Not Yet Posture Marco Rabuffetti, Gabriele Bovi, Pier Luigi Quadri, Davide Cattaneo, Francesco Benvenuti, and Maurizio Ferrarin, Member, IEEE

Abstract—A ground reaction based method is proposed to evaluate the hypothesis that a stabilization phase occurs in transitions towards erect posture, following the macroscopic movement and preceding the quiet final erect posture, whose aim is to control and dissipate the residual inertial unbalancing forces occurring at the transition end. The experimental protocol considers three tasks leading to the final erect posture: taking a step forward (F), sit-to-stand (S), and bending the trunk forward (B), The method mainly consists of the fitting of a negative exponential function on the instability time profile following the end of the transition movement. The model parameters 0 , , and inf , respectively, quantify the initial instability rate, a time duration related to the stabilization, and the final asymptotic instability rate. Results from a sample of 40 adult able bodied subjects demonstrated that a postural stabilization phase actually occurs: inf is smaller (0.010, 2 for, respectively, , , and tasks) than 0.010, and 0.008 0 2 ). Tis in the order of seconds (0.95, (0.081, 0.137, and 0.057 0.51, and 1.00 s). No trial with large values of both 0 and was observed, evidencing that large initial instability rates are quickly controlled and reduced. The 0 and parameters distribution are discussed according to the possible underlying active and/or passive stabilization mechanisms. The test–retest reliability overall figure (mean ICC 0.45 for 12 indexes) increased, when dropping the indexes related to the less reliable B task, to values (mean ICC 0.56 for eight indexes) comparable to published posturographic data.



Index Terms—Fall risk, ground reaction, human movement, motor control, posture.



OSTURES, and particularly quiet erect posture, actually consist of small oscillations of the body whose continuously changing acceleration reflects on the ground reaction force components. Though pathologies and ageing are known to affect the performance in postural tasks, the relation among outcomes of instrumental posturographic tests and fall risk is Manuscript received November 03, 2010; revised March 30, 2011; accepted May 21, 2011. Date of publication June 23, 2011; date of current version August 10, 2011. This work was supported in part by Ricerca Corrente of the Italian Ministry of Health, in part by Fondazione Telethon—UILDM (Grant GUP10010), and in part by the “Fondazione per lo Studio delle Malattie Neurodegenerative delle Persone Adulte e dell’Anziano,” Lugano, Switzerland. M. Rabuffetti is with the Fondazione Don Carlo Gnocchi ONLUS, Polo Tecnologico, I-20148 Milano, Italy (e-mail: [email protected]). G. Bovi, D. Cattaneo, and M. Ferrarin are with the Fondazione Don Carlo Gnocchi ONLUS, I-20148 Milano, Italy. P. L. Quadri is with the Ospedale Beata Vergine, CH-6850 Mendrisio, Switzerland. F. Benvenuti is with the AUSL 11, I-50053 Empoli, Italy. Digital Object Identifier 10.1109/TNSRE.2011.2159241

questionable [1] and strong evidences have been produced that an history of fall and a simple clinical examination are more efficient fall predictors than the outcome of instrumental balance assessments [2]. Transitions to or from postures frequently occurring in everyday life (such as, for example, initiation or termination of gait, sit-to-stand or stand-to-sit, etc.) require the modulation and switch between specific motor programs [3]; “transfer dysfunctions” have been included among relevant fall risk factors [4] and, interestingly, these tasks are earlier affected by Parkinson disease while cyclic movements and postures are less affected [5], [6]. Particularly, in transitions toward an erect posture, such as in gait termination, the key point is “how the nervous system anticipates, controls and arrests forward momentum of the body” [7] in order to avoid excessive unbalancing inertial components still to be controlled at the end of the transition. Consequently, transitions towards an erect posture, which healthy people can accomplish very easily and safely, may represent a potential factor of falling risk for frailer individuals as observational and experimental studies have evidenced. Nyberg [8] reported that out of a total of 153 falls recorded consecutively in a geriatric stroke rehabilitation unit, 37% occurred while transferring between postures, 20% when negotiating a sitting posture and 15% while walking. Nylsagard [9] reported 270 falls in 48 out of 76 MS patients under screening and most of the falls occurred indoors during daily activities. Robinovitch [10] reported in an observational study that, out of 81 video-recorded falls from community-dwelling older people, 32% were caused by an incorrect body weight transfer, 28% occurred during standing, 15% during forward stepping, 14% while initiating walking. Najafi [11] studied sit-to-stand transitions in frail elderly subject, using accelerometers and gyroscopes based on micro electro-mechanical systems (MEMS), evidencing that some performance indexes, such as the duration of the transitional movement or the number of attempts before succeeding, are predictors for fall risk. While extensive literature is available on transitions, comparatively few studies focused on what happens when a transitional movement ends and posture is assumed to be established: Stelmach [12] reported the occurrence of a “recovery phase” following an upper-limb movement in which erect posture was obtained gradually—in particular he showed that a one-second sway area index progressively tended to steady state values, and that postural recovery duration was prolonged in aging subjects if dual tasks were involved, thus implying an attentional load;

1534-4320/$26.00 © 2011 IEEE


Mazzà [13] introduced the concept of “preparation to the upright posture” following a sit-to-stand movement and preceding the final erect posture—she studied 22 young, able-bodied subjects and observed that the duration of the postural preparation phase increased when feet distance decreased, thus evidencing how a mechanical factor such as the reduction of the base of support area could make the task more difficult. The present study focuses on the motor behavior following the end of a transition (i.e., a macroscopic movement): when it is generally assumed that the standing posture is promptly reached, we conversely hypothesize the occurrence of a phase of “postural stabilization” progressively leading to a steady state erect posture. In our hypothesis, the postural stabilization aims at the control and dissipation of residual inertial unbalancing components, i.e., the residual momentum. In this view, the postural stabilization is no more a transitional movement because it is characterized by body oscillations reflected by center of mass (COM) horizontal acceleration which continuously changes sign, and, at the same time, it is not a steady state posture since the amplitude of such oscillations, and related accelerations, is still changing (decreasing). In simple words, it is no more a movement, not yet a posture. The concept of postural stabilization has been already considered, explicitly or implicitly, in scientific literature: the identification of a staggering behavior in response to external small perturbations was reported [14], [15], the damping component involved in a passive model of ankle control of posture [16] implicitly supported a progressive dissipation of excessive body momentum; it was evidenced that posture is not a stationary phenomenon but may alternate phases with changing postural performance indexes [17]; in activities requiring great motor skill such as landing from a jump, it was described a stabilization phase following landing which can be characterized by duration and amount of sway [18]. The latter citation particularly evidences how transitional movements towards an erect posture, though programmed and controlled in order to end with the posture itself, generally show a further stabilization after their conclusion before a steady-state posture is reached. Based on the above considerations, the rationale for studying the postural stabilization is that subjects, particularly those with general frailty and/or with motor control impairments, when reaching an erect posture are challenged by the need to dissipate the residual instability resulting from the transfer movement. A quantification of indexes, related both to the instability levels and to the temporal aspects of the stabilization, provides a detailed profile of the subjective stabilizing ability. This assessment, when performed in several daily life movements, may be related to the risk of falling, particularly in elderly and in patients with motor disorders, and, more in general, it is complementary to balance assessment done by classic instrumental posturography during quiet erect posture. The aims of this study are to evidence the occurrence of a postural stabilization following a transient movement and preceding a steady state erect posture and to quantify indexes related to the postural stabilization performance. The novelty of this study consists in the proposal of a mathematical model with an exponential decay factor, whose independent parameters, related to center of mass (COM) acceleration and not to the center


of pressure (COP) trajectory, quantify various aspects about the postural stabilization performance. The model was applied to a large experimental data set obtained from 40 healthy subjects and including three transient movements, chosen because of their different features but nonetheless belonging to the repertoire of daily activities, leading to erect posture. This approach allows to characterize the stabilization performance in relation to the amount of initial and final instability, going beyond the limitation of descriptions relying on a single index (generally the duration of stabilization). II. MATERIALS AND METHODS A. Instrumentation, Protocol and Subjects The tasks included in the protocol were performed with eyes open and characterized as follows. : . The subject sat quiet on a chair next to 1) Sit-to-Stand the force plate, keeping both feet on the plate and aligned with the inter anterior-superior iliac spine distance. The subject was asked to stand up with usual speed, without any movement of the upper limbs or the feet, and then to keep upright balance as still as possible for 10 s while looking at a target at eye level, about one meter distant. The rationale of this task was to study a transition requiring significant lower limb strength exertion. : The subject stood next to 2) Taking a Step Forward the force plate, looking at the visual target located in front of him. When asked to by the operator, he stepped on the plate and stopped, standing as still as possible for 10 s. The feet position was not standardized, though we instructed the subject not to move them once set down. The rationale of the task was to simulate balance perturbation following step-size transfer at slow speed with less strength exertion than sit-to-stand. : The subject stood still 3) Bending the Trunk Forward on the force plate, with the feet positioned like in task S, looking at the visual target at eye level. When asked to by the operator, the subject bent the trunk in order to reach and touch with the right hand a target positioned 0.1 m in front of his left knee, and then returned to the initial position standing as still as possible for 10 s. The rationale for this task was to study the stabilization following a movement implying a relatively large strength at hip-lumbar level, an attentional load involved in the reaching-related eye-hand coordination, and a stimulation of the vestibular system. The experimental protocol required to perform each task for three times in a predefined sequence (S/F/B), for a total of nine trials. This number of repetitions was set according to previous studies [19], [20], and we consider the total time of about 20 minutes, needed to perform the whole experiment, compatible with frailer individuals (such protocol is expected to be applied to elderly, to patients with severe neurodegenerative diseases or suffering from brain lesions) who are likely to experience a fatigue effect. The global dynamic associated to task performance (ground reaction) was measured by a piezoelectric force platform (Kistler, CH) with a sampling rate of 960 Hz. The recruited group consisted of 40 healthy subjects (20 ) males and 20 females, aged from 17 to 79 (mean years). All participants provided informed consent.




Fig. 1. Vertical and anterior–posterior components of ground reaction force (GRFv and ) plotted versus time in the three considered tasks (S for sit-tostand, F for taking a step forward, B for bending the trunk forward). Time t , marking the beginning of the postural stabilization phase, is evidenced. For a better is magnified by a factor 3. A concise description of the algorithms for t identification [derived from 21] follows; tasks S and B: 1) maximum visualization, vertical force peak is identified (corresponding to maximum vertical inertia); 2) following minimum force peak is identified (corresponding to minimum vertical inertia); 3) t is defined as the first sample higher than body weight; task F: 1) maximum vertical force peak is identified (corresponding to hind foot contact on the force plate); 2) t is defined as the first sample lower than body weight.


B. Data Analysis: Identification of Key Events and Definition of the Instability Profile of The key event in postural stabilization is the time ending of macroscopic focal transition movement and consequent starting of the postural stabilization phase. We determined on the vertical component of GRF by identification algorithms [21] for each task (see Fig. 1). Since the tasks we propose mainly develop in the subject’s sagittal plane, the profile of the anterior–posterior component , obtained by dividing by of COM acceleration ( body-mass) is related to the instability of the subject. was first low-pass filtered (Butterworth The variable fifth-order low-pass filter, 3-Hz cutoff frequency), then the root was computed in a 1 s window (epoch) mean square of moving forth from , and each computed value was associated to the corresponding epoch initial instant [12]. The resulting accounts for variability. time series Finally, an instability time profile was identified as the median series obtained from repeated executions, profile from synchronized on (see Fig. 2). C. Data Analysis: Stabilization Model and Related Parameters The commonly observed profile showed a decrease in amplitude from on, initially at a fast rate and progressively getting flat. A negative exponential mathematical model [see profile (1)] was adopted to fit the (1) The fitting of the model on the experimental data is fully described by three parameters , , and , where . Parameter accounts for the residual instability at the end of focal movement (or, alternatively, at the beginning of the is the corresponding stabilization phase), while parameter asymptotic value; both parameters have the same physical di. mensions of Parameter , namely the inverse of the decay rate as derived from the tangent at , is dimensionally a time and is proportional to the time needed to stabilize (about three times is

Fig. 2. Example of the fitting of the negative exponential model (thick gray curve as obtained from three repetitions line) on the reference median (thin black lines) of task S . Values of the model parameters, representing postural stabilization indexes, Y , T , and Y are graphically evidenced.


generally assumed as the time lag required to pass from to ). . Finally it was defined an integral parameter Due to the inherent not normal distribution of the indexes (always positive and generally close to zero), nonparametric descriptors and statistical tests were adopted. D. Inter-Session Reliability In order to assess the test–retest reliability of the listed parameters, we compared measurements from two sessions, about one week apart, on a subgroup of 10 subjects. Additionally, in order to evaluate the effect of the number of tasks repetitions on reliability, we asked each subject to perform five additional repetitions for each task. The inter-session reliability was quantified for the standard protocol, and for the extended data set including additional trials, by intra-class correlation coefficients, model 2,k, and standard error of measurement (SEM) [22].




( = 40)

Fig. 3. Percentages of detected body oscillations in 1 s epochs, one epoch preceding and ten epochs following t , in each task. Body oscillation is detected changes sign at least once within the epoch. N when for each task.


= 120

III. RESULTS resulting from the 360 performed The analysis of the measurements (40 subjects, three tasks, three repetitions per task) initially implied the identification of , final instant of the transfer movement and initial instant for the oscillatory state typical of erect posture. The detection of sign changes of (a landmark of body oscillation in the anterior–posterior direction) in epochs lasting 1 s, lets to compute the percentages of oscillating state in each epoch for each task such as reported in Fig. 3: in the first 1 s epoch following percentage is 93.0 for the task (rising up to 96.5 in the second epoch) and 97.4 for the S task, then globally tending to the ceiling value of 100%. The percentage of oscillating state in the epoch preceding is larger in those tasks involving a vertical displacement, particularly task which is a on-the-spot task, task trials show oscillations which while about half of the are to be classified as belonging to the transition movement. After pooling the three repeated measurements synchronized proat , the model of stabilization was applied to 120 files (40 subjects, three tasks) and the stabilization parameters computed. The 5th, 50th, i.e., the median, and 95th percentiles , and , are reported in Table I. values of parameters , , Statistically significant differences across tasks (Kruskall Wallis nonparametric analysis of variance and related post-hoc , and : with a p-value of 0.05) were found for , was larger than in all tasks; • was larger (about doubled) in task than in task and • ; • was shorter (about a half) in task compared to task and task ; was lower in task with respect to task . • No significant differences across tasks were found for param. eter and , In order to investigate the relation between a scatter-plot of those parameters is presented in Fig. 4. A perusal of the plots lets to adopt a hyperbolic model

( = 40) =

Fig. 4. Scatter-plots N of T versus Y , for the considered tasks (S for sit-to-stand, F for taking a step forward, B for bending the trunk forward). An hyperbolic model (Y 1 T I where I : = ) is applied in order to identify the 95% confidence upper interval limit (solid line).

= 0 109 m s

to identify a limit curve including 95% of the sample. and SEM for The results of the test–retest analysis, all parameters, for the subgroup of 10 subjects, are presented in Table II. , As to considered tasks, ICCs were satisfactory ( in tasks and ; task according to [20]) for , and had generally unsatisfactory ICC, except for . As to computed parameters, parameter I showed unsatisfacshowed the best values. tory reliability across tasks, while Parameters and showed satisfactory values for tasks and , but not for task . The general figure about test–retest reliability was influenced when increasing the number of repetitions from 3 to 8: the ICC overall figure improved from an average of 0.45 (sd 0.31)





to an average of 0.61 (SD 0.19). When excluding the task B, which presented unsatisfactory test–retest performance, the ICC overall figure for three repetitions per task was 0.56 (sd 0.14). IV. DISCUSSION The tasks we selected were easy to perform and to understand, and involved different combinations of motor strategies and sensory inputs. Comparatively, task F (taking a step forward) was the least demanding in term of muscular effort and visuo-vestibular coordination, task S (sit-to-stand) required high muscular recruitment of hip and knee extensors while task B (bending the trunk forward) was the most demanding for the integration of visuo-vestibular information due to large movement of the head. All tasks involved displacements of body masses in order to introduce relevant inertial components to be managed in the performance. Nonetheless, the tasks’ performance was entirely controlled by the subject and no commitment was made about speed of transfer, particularly the amount of instability at the beginning of the stabilization phase was not controlled as in experimental setups studying the response to external predefined perturbations [23]–[26]. The single session, consisting of nine trials, required in average less than 20 min for each subject. The identification of , the instant in which the transfer movement ends and the stabilization phase is supposed to begin, generally corresponded to the onset of an oscillatory behavior marked by changes of sign of the filtered anterior–posterior GRF component within 1-s epochs. In a small percentage (with a maximum of 7% in the first 1-s epoch after of the F task, and generally amounting to about 1%–2%) no sign change was observed, which we interpret as the occurrence of a particularly slow oscillation, for example a slow transfer of COP between the rear-foot and the fore-foot which may take more than one second. Conversely, the occurrence of sign changes of in the epoch preceding is explained by the mostly-vertical displacement in and tasks and by accommodations occurring during the end of the monopodalic phase in task. These results confirmed the opportunity to consider the vertical GRF

component in identifying the end of the transfer movement [21], particularly when the movements imply large vertical displacements. The definition of indexes was based on the COM A/P acceleration time profile and not on COP trajectory features [12], [13], because this balance-related variable is characterized by oscillation around a fixed value, the null value, while the COP trajectory may be characterized, besides back-and-forth displacements, by a slow wandering behavior (that cannot be properly considered an oscillation) inside the base of support. The presented results confirmed our hypothesis on the existence of a postural stabilization phase: given an oscillatory bewas havior, in all considered tasks, the final instability rate lower, about one magnitude order, than the initial instability occurring at the end of the transfer movement. Consequently, the postural stabilization consists of negotiating a certain amount of at the end of a transfer movement in a finite time instability duration (proportional to ) in order to reach the final steady instability). state erect posture (characterized by The initial instability is expected to be influenced by the transfer movement just performed, particularly the amount of AP displacement involved and the transfer speed. In difficult and demanding transitory tasks, also the level of motor skill has [18]; however this influence is not exshown to influence pected in easier common-life tasks, such as the ones considered in the present study. The presented results show that initial instability was characterized by a relatively wide range of values: though there was a tendency to smaller values close to the inferior limit, larger values also occur. The mean time needed to stabilize, which can be assumed to be three times (this choice identifies the time needed to dissipate about 95% of the initial instability, being the mathematical expression of the remaining fraction of the initial value at time expressed as ), for the healthy population was 1.5, 3.0, and 2.8 s, respectively, for task , , and . The latter value of 2.8 s (sd 1.0 s) is directly comparable with the value 2.7 s (sd 0.8 s) of the “duration of the preparation to upright posture” as reported by Mazzà [13] for a sit-to-stand movement. More generally, the reported values of duration of the stabilization phase are comparable, though smaller, with the “recovery time” reported by Stelmach [12] after a self-perturbation consisting in several fast arm swings and additionally including a motor or mental dual-task: in his experiments, young people took from 4 to 6 s, while elderly were significantly slower (up to 11 s); the different outcomes of the two studies can be explained with the different tasks preceding the stabilization and/or by the different indexes’ definitions, nonetheless both approaches evidenced that recovery/stabilization may require a time in the order of some seconds. Parameter , accounting for the stabilization duration, should and to the skill of the be related to the initial instability performer. In the present study, speculations are possible only about the first factor, being the experimental subjects sampled from an able-bodied adult population. As to the possible relation and , the most important evidence is that no coubetween pling was observed (see Fig. 3) between large values of and large values of . Though this evidence cannot support the identification of a strict inverse relationship between the two param-


eters, an hyperbolic function can be identified as the upper limit of the distribution of and . Since upright balance is believed to be maintained through a mix of active [27], [28] and passive [16], [29] mechanisms, which can be eventually activated in an intermittent mode [30], and the switching across different active motor programmes is modulated by speed in gait terminaand , in tion [31], our hypothesis about the distribution of which the upper limit for is apparently modulated by , is that different mixes of active and passive mechanisms, involving different response timing, may occur depending on the amount of initial instability. In this scenario, a degradation of such stabilizing mechanisms, possibly occurring in subjects with motor disorders, could result in a combination of and beyond the limiting function, which means that a longer than normal time would be required to stabilize a certain instability. On the contrary, it can be argued that the proposed experimental paradigm may produce false negative results when considering motor-impaired subjects if initial instability were small, and therefore not challenging balance, due to a slower than normal transitional movements. Such remarks make it reasonable to carry on future researches about postural stabilization in selected populations of patients with sensory-motor impairments. The discussion about initial instability, which is not controlled in the proposed method, is further evoked when considering that most of the experimental studies on perturbed balance involved, on the contrary, known, at least by the experimenter, perturbations challenging the balance control system [23], [26], if not producing a loss of balance and a subsequent forward stepping [24], [25]. Though the proposed method can support also the study of postural stabilization following controlled perturbations, we consider that it is worth studying the postural stabilization “as is” following a transfer movement. and , whose main The study of the interplay between feature consists of the limiting hyperbolic curve, supported the and : large definition of the parameter I as the product of parameter I values correspond to large and , while small parameter I values may correspond either to high rapidly stabilized (small ) or to small with prolonged stabilization (possibly large ). Future studies on populations including fallers and not fallers will address the hypothesis that parameter I might represent a fall risk predictor. is expected to be The final asymptotical instability related to erect posture only, being elapsed a sufficiently long time since whatever transitional movement have been were observed in the task charperformed: while larger acterized by the largest body horizontal displacement, showed no difference among tasks. The latter observation implicitly guarantees the validity of posturographic tests given that a sufficient time elapsed since positioning on the platform. The assessment of test–retest reliability on a subgroup of 10 and large SEM values. subjects produced relatively low Particularly overall figure showed an average value of 0.45 (sd 0.31), where published data showed, in the assessment of reliability of a posturographic test on geriatric patients, an figure, computed on 13 parameters, of 0.67 (sd overall 0.18) [32]. The obtained from a eight-trials version (instead of the three-trials setting) of our test (ICC mean 0.61, sd 0.19) was much closer to the cited reliability data, nonetheless


the improvement in reliability was not considered worth of a triplication of experimental time particularly in the assessment of easily fatigable subjects, such as elderly and patients, theredata fore the number of repetitions was set to 3. Low for task B suggest to drop it from future clinical applications, despite its strong rationale. If excluding task B, the overall figure had, for the chosen three-trials setting, an average value of 0.56 (sd 0.14). The SEM values reported in Table I let to compute the minimally detectable differences, i.e., the change that could be considered clinically different between two measurements because greater than the noise induced by the intrinsic variability of the phenomena and instrumentation [19]. The presented test–retest reliability characteristics of the postural stabilization assessment, compared with classic posturography, could warn against its application. However it should be remarked that postural stabilization assessment focuses on different tasks which may be differently affected than steady state posture by balance disorders. Similarly it has been shown that transitory locomotor functions, such as turning gait, are more sensitive experimental paradigms to the underlying pathology than steady state rectilinear gait [33]. From a methodological point of view, one of the interesting features of the model here presented is the possibility to be applied to different transitional tasks. This allows an extensive assessment of the subcomponents of the balance system giving a more detailed view on the subject’s balance skills. Although the present study focused on self-induced uncontrolled perturbations, the method here presented allows to study also the postural stabilization following external and/or controlled perturbations [25], [26] (in this latter case is fixed and the perturbation is predefined). inducing Furthermore, it is worth to evidence the focus on force components rather than COP components, the latter being already adopted in several papers [13], [18]: from a methodological point of view the two approaches are substantially equivalent and both can be easily managed when adopting a dynamometric force platform as experimental setup, but the added value of our force(acceleration)-based approach consists of its implementability on MEMS accelerometric wearable setups, which allow for feasible outside-the-lab applications of postural assessment and rehabilitation [34], [35]. REFERENCES [1] M. Piirtola and P. Era, “Force platform measurements as predictors of falls among older people—A review,” Gerontology, vol. 52, pp. 1–16, 2006. [2] P. Gerdhem, K. A. Ringsberg, K. Akesson, and K. J. Obrant, “Clinical history and biologic age predicted falls better than objective functional tests,” J. Clin. Epidemiol., vol. 58, no. 3, pp. 226–232, 2005. [3] P. Crenna, D. M. Cuong, and Y. Brénière, “Motor programmes for the termination of gait in humans: Organisation and velocity-dependent adaptation,” J. Physiol., vol. 537, pp. 1059–1072, 2001. [4] M. E. Tinetti, D. I. Baker, P. A. Garrett, M. Gottschalk, M. L. Koch, and R. I. Horwitz, “Yale FICSIT: Risk factor abatement strategy for fall prevention,” J. Am. Geriatr. Soc., vol. 41, pp. 315–320, 1993. [5] P. Crenna, I. Carpinella, M. Rabuffetti, E. Calabrese, P. Mazzoleni, R. Nemni, and M. Ferrarin, “The association between impaired turning and normal straight walking in Parkinson’s disease,” Gait Posture, vol. 26, pp. 172–178, 2007. [6] I. Carpinella, P. Crenna, E. Calabrese, M. Rabuffetti, P. Mazzoleni, R. Nemni, and M. Ferrarin, “Locomotor function in the early stage of Parkinson’s disease,” IEEE Trans. Neural Syst. Rehabil. Eng., vol. 15, no. 4, pp. 543–551, Dec. 2007.



[7] S. D. Perry, L. C. Santos, and A. E. Patla, “Contribution of vision and cutaneous sensation to the control of centre of mass (COM) during gait termination,” Brain Res., vol. 913, no. 1, pp. 27–34, 2001. [8] L. Nyberg and Y. Gustafson, “Patient falls in stroke rehabilitation. A challenge to rehabilitation strategies,” Stroke, vol. 26, pp. 838–842, 1995. [9] Y. Nilsagard, C. Lundholm, E. Denison, and L. G. Gunnarsson, “Predicting accidental falls in people with multiple sclerosis: A longitudinal study,” Clin.Rehabil., vol. 23, pp. 259–269, 2009. [10] S. Robinovitch, F. Feldman, D. Wan, O. Aziz, and T. Sarraf, “Video recording of real-life falls in long term care provides new insight on the cause and circumstances of falls in older adults,” in Proc. XIX Conf. ISPGR, 2009, pp. 80–81. [11] B. Najafi, K. Aminian, F. Loew, Y. Blanc, and P. A. Robert, “Measurement of stand-sit and sit-stand transitions using a miniature gyroscope and its application in fall risk evaluation in the elderly,” IEEE Trans. Biomed. Eng., vol. 49, no. 8, pp. 843–851, Aug. 2002. [12] G. E. Stelmach, H. N. Zelaznik, and D. Lowe, “The influence of aging and attentional demands on recovery from postural instability,” Aging, vol. 2, pp. 155–161, 1990. [13] C. Mazzà, M. Zok, and U. D. Croce, “Sequencing sit-to-stand and upright posture for mobility limitation assessment: Determination of the timing of the task phases from force platform data,” Gait Posture, vol. 21, pp. 425–431, 2005. [14] L. M. Nashner, “Fixed patterns of rapid postural responses among leg muscles during stance,” Exp. Brain Res., vol. 30, pp. 13–24, 1977. [15] M. E. Tinetti, “Performance-oriented assessment of mobility problems in elderly patients,” JAGS, vol. 34, pp. 119–126, 1986. [16] D. A. Winter, A. E. Patla, F. Prince, M. Ishac, and K. Gielo-Perczak, “Stiffness control of balance in quiet standing,” J. Neurophysiol., vol. 80, pp. 1211–1221, 1998. [17] A. Bottaro, M. Casadio, P. G. Morasso, and V. Sanguineti, “Body sway during quiet standing: Is it the residual chattering of an intermittent stabilization process,” Hum. Movement Sci., vol. 24, pp. 588–615, 2005. [18] P. McKinley and A. Pedotti, “Motor strategies in landing from a jump: The role of skill in task execution,” Exp. Brain Res., vol. 90, no. 2, pp. 427–440, 1992. [19] D. Lafond, H. Corriveau, R. Hebert, and F. Prince, “Intrasession reliability of center of pressure measures of postural steadiness in healthy elderly people,” Arch. Phys. Med. Rehabil., vol. 85, pp. 896–901, 2004. [20] T. L. Doyle, R. U. Newton, and A. F. Burnett, “Reliability of traditional and fractal dimension measures of quiet stance center of pressure in young, healthy people,” Arch. Phys. Med. Rehabil., vol. 86, pp. 2034–2040, 2005. [21] B. Etnyre and D. Q. Thomas, “Event standardization of sit-to-stand movements,” Phys. Therapy, vol. 87, pp. 12–12, 2007. [22] P. E. Shrout and J. L. Fleiss, “Intraclass correlations: Uses in assessing rater reliability,” Psychol. Bull., vol. 86, no. 2, pp. 420–428, 1979. [23] M. L. Latash, A. S. Aruin, I. Neyman, and J. J. Nicholas, “Anticipatory postural adjustments during self inflicted and predictable perturbations in Parkinson’s disease,” J. Neurol. Neurosurg. Psychiatry, vol. 58, no. 3, pp. 326–334, Mar. 1995. [24] M. C. Do, C. Schneider, and R. K. Chong, “Factors influencing the quick onset of stepping following postural perturbation,” J. Biomech., vol. 32, no. 8, pp. 795–802, Aug. 1999. [25] C. J. Hasson, R. E. Van Emmerik, and G. E. Caldwell, “Predicting dynamic postural instability using center of mass time-to-contact information,” J. Biomech., vol. 41, no. 10, pp. 2121–2129, Jul. 2008. [26] H. Trivedi, J. A. Leonard, L. H. Ting, and P. J. Stapley, “Postural responses to unexpected perturbations of balance during reaching,” Exp. Brain Res., vol. 202, no. 2, pp. 485–491, 2010. [27] P. G. Morasso and M. Schieppati, “Can muscle stiffness alone stabilize upright standing,” J. Neurophysiol., vol. 83, pp. 1622–1626, 1999. [28] P. G. Morasso and V. Sanguineti, “Ankle stiffness alone cannot stabilize upright standing,” J. Neurophysiol., vol. 88, pp. 2157–2162, 2002. [29] D. A. Winter, A. E. Patla, F. Prince, and M. G. Ishac, “Ankle muscle stiffness in the control of balance during quiet standing,” J. Neurophysiol., vol. 85, pp. 2630–2633, 2001. [30] A. Bottaro, Y. Yasutake, T. Nomura, M. Casadio, and P. Morasso, “Bounded stability of the quiet standing posture: An intermittent control model,” Hum. Mov. Sci., vol. 27, pp. 473–495, 2008.

[31] P. Crenna, D. M. Cuong, and Y. Brénière, “Motor programmes for the termination of gait in humans: Organisation and velocity-dependent adaptation,” J. Physiol., vol. 537, no. 3, pp. 1059–1072, 2001. [32] F. Benvenuti, R. Mecacci, I. Gineprari, S. Bandinelli, E. Benvenuti, L. Ferrucci, A. Baroni, M. Rabuffetti, M. Hallett, J. M. Dambrosia, and S. J. Stanhope, “Kinematic characteristics of standing disequilibrium: Reliability and validity of a posturographic protocol,” Arch. Phys. Med. Rehabil., vol. 80, no. 3, pp. 278–287, 1999. [33] I. Carpinella, P. Crenna, E. Calabrese, M. Rabuffetti, P. Mazzoleni, R. Nemni, and M. Ferrarin, “Locomotor function in the early stage of Parkinson’s disease,” IEEE Trans. Neural Syst. Rehabil. Eng., vol. 15, no. 4, pp. 543–551, Dec. 2007. [34] L. Chiari, M. Dozza, A. Cappello, F. B. Horak, V. Macellari, and D. Giansanti, “Audio-biofeedback for balance improvement: An accelerometry-based system,” IEEE Trans. Biomed. Eng., vol. 52, no. 12, pp. 2108–2111, Dec. 2005. [35] D. Giansanti, “Investigation of fall-risk using a wearable device with accelerometers and rate gyroscopes,” Physiol. Meas., vol. 27, no. 11, pp. 1081–1090, Jul. 2006.

Marco Rabuffetti received the M.S. degree in electronic engineering from the Politecnico of Milano, Milan, in 1991. He spent the year 1990 at the Mario Negri Institute for Pharmacological Researches in Milano. Since 1991, he worked as a researcher in the biomedical field: from 1991 to 2003 at the Centro di Bioingegneria of the Politecnico of Milano and of the Fondazione Don Gnocchi, since 2003 up to present days as a Senior Researcher at the Polo Tecnologico of the Fondazione Don Gnocchi in Milano. Since 2001 he has been in charge of academic courses for the Politecnico of Milano, Bioengineering Department (2002-2003) and for the Università degli Studi di Milano, Faculty of Medicine (2001-present). His main interests include biomechanics and motor control of human movement, modeling of cognitive functions. He is coauthor of more than 40 papers on peer-reviewed scientific journals including IEEE Transactions, Brain, Journal of Biomechanics, Medical & Biological Engineering & Computing, and Gait & Posture. Mr. Rabuffetti was founder member of the Italian Society for Movement Analysis in Clinic (SIAMOC) in 1999 and served in the Society Board.

Gabriele Bovi, photograph and biography not available at the time of publication.

Pier Luigi Quadri, photograph and biography not available at the time of publication.

Davide Cattaneo, photograph and biography not available at the time of publication.

Francesco Benvenuti, photograph and biography not available at the time of publication.

Maurizio Ferrarin, photograph and biography not available at the time of publication.

Lihat lebih banyak...


Copyright © 2017 DADOSPDF Inc.