Exploring Sway Metrics of Intermittency in the Feedback Postural Control for Assessing Balance Impairment in Patients with Parkinson's Disease

Kazuki Matsui, Yasuyuki Suzuki, Akihiro Nakamura, Charles E. Smith, Kunihiko Nomura, Takuyuki Endo, Saburo Sakoda, Pietro G. Morasso, Taishin Nomura

Abstract: Postural control during quiet stance is characterized by the center of pressure (CoP) time series. Recently, we showed that CoP data from healthy young (HY) and elderly (HE) individuals, and patients with Parkinson’s disease (PD) can be well fitted by the intermittent control model with different parameter values representing the degree of intermittency in the feedback control. We performed a model-based classification (MBC) to show that the subjects could be classified into intermittent control (INT) group including all HYs, active (CONT) and passive continuous control (STIFF) groups including HE and PD. Here, we showed first that the sum of gait-posture-related UPDRS scores for the CONT and STIFF was significantly higher than that for the INT, indicating that the MBC is a good biomarker for assessing the balance impairment in PD. We then quantified CoP data from each individual using a variety of sway metrics, and explored metrics that have a classification capability equivalent to the MBC. Multiple metrics were identified as such metrics, suggesting that the MBC-based assessment can be achieved using those metrics without performing the computationally heavy model fitting.