Gait and Posture 10 (1999) 129–134 Instantaneous postural stability characterization using time-frequency analysis Mohammed Ferdjallah a,b , Gerald F. Harris b, *, Jacqueline J. Wertsch c a Orthopaedic and Rehabilitation Engineering Center (OREC), Marquette Uniersity, 1515 W. Wisconsin Ae., Box 25, Milwaukee, WI 53203, USA b Shriners Hospital for Sick Children, 2211 N Oak Park Ae., Chicago, IL 60635, USA c Department of Physical Medicine and Rehabilitation, Medical College of Wisconsin, Milwaukee, WI 53226, USA Abstract Postural stability assessment is critical to a more accurate understanding of sway and balance control. The center of pressure (COP) metric has been shown to be a suitable output measure for time and frequency analysis. However, the center of pressure is a non-stationary signal. Standard time and frequency analysis methods may not be adequate for monitoring the dynamic changes in the center of pressure signal. In this study a time-frequency method, based on data-adaptive evolutionary spectral estimation, is applied to monitor the dynamic changes of the center of pressure in a non-stationary environment. Metrics including the instantaneous mean frequency (IMF), instantaneous spectral bandwidth (ISB), and instantaneous average power (IAP) are analyzed to characterize the center of pressure signal in both the anterior-posterior (AP) and the medial-lateral (ML) planes. Within the confines of this study, the IMF was found to be inversely proportional to IAP. The inverse proportionality factors were calculated in both eyes-open and eyes-closed trials during upright quiet standing. These findings suggest that the time-frequency analysis provides instantaneous metrics which describe the amplitude changes and frequency shift of the center of pressure under a variety of environmental conditions, thus providing a more reliable quantification of postural control. © 1999 Published by Elsevier Science B.V. All rights reserved. Keywords: Postural control; Balance; Center of pressure (COP); Time-frequency analysis; Instantaneous mean frequency (IMF); Instantaneous spectral bandwidth (ISP); Instantaneous average power (IAP) www.elsevier.com/locate/gaitpost 1. Introduction Stable control of posture and balance is an important goal for vestibularly impaired patients, children with cerebral palsy, and others lacking stability due to pathol- ogy or injury. For instance, children with cerebral palsy frequently have an impaired sense of equilibrium, abnor- mal motor control, persistence of primitive reflexes, and may develop abnormal posturing. Clinical assessment of balance control is therefore critical and often requires integrated functional, systems and quantitative ap- proaches [1,2]. Means to assess posture by constructing metrics which reliably identify stable control have long been a focus of investigators [3–6,18]. Using standard force platforms, numerous authors have analyzed center of pressure data to quantitatively describe postural sway characteristics. The center of pressure represents the global position of the ground reaction force under the supporting surface of a subject’s feet. Thus, the center of pressure signal is an indirect measure of the body sway. Various sampling frequencies and durations have been employed to better characterize the center of pressure. Under a variety of protocols designed to assist the subject in maintaining stable control, the center of pressure in both anterior-posterior (AP) and medial-lateral (ML) planes has proven to be a significant output metric [5]. The path length, sway area and AP sway ranges have also been shown to be effective parameters for monitoring postural sway [5]. Standard analyses including time and frequency methods are extensively reported in the literature to characterize the center of pressure. Fourier analysis has been used to compare the amplitude spectra of the components of sway within multiple frequency bands [7]. Fifth order autoregressive models have been developed to character- ize the autoregressive structure of postural sway [2,3]. * Corresponding author. 0966-6362/99/$ - see front matter © 1999 Published by Elsevier Science B.V. All rights reserved. PII:S0966-6362(99)00023-5