Unobtrusive Assessment of Bipedal Balance Performance

Reduced postural stability are symptoms of many medical conditions. Depending on the conditions, there are training or medication strategies to ameliorate the balance of a subject. After medical treatment (surgery etc.) or in the context of an intervention/recovery strategy, these patients are often assessed in functional gait assessments (FGA). An expert, e.g. a Physio Therapist, can decide based on the outcome of such an assessment on future treatment plans. FGAs are often performed with no technological assistance: a subject performs pre-dened tasks and the performance is evaluated visually by an expert. Existing technological assessment tools are scarcely used due to time and monetary restrictions. In this paper, we present a wearable sensor system that can be used for FGAs. Our system comprises a pressure-sensing component and inertial sensors to assess features known to correlate with balance. We validated our system against technological state of the art. We used the system on 6 patients and 5 healthy subjects. The system can distinguish between normal stance and stance with reduced postural control with an accuracy of more than 93%. Walk- ing episodes were classied into two categories with 91%. Based on features of stance and features of gait, our system can discriminate between healthy subjects and subjects with reduced postural stability with an accuracy of 94%