Examining Wireless Accelerometer Filtering Cutoff Frequencies In Measuring Peak Ground Reaction Force During Functional Tasks

Force plates are considered the gold standard when measuring peak ground reaction force (pGRF), however, the widespread clinical and sport application may be limited due to their expense and lack of portability. Wearable sensors, such as wireless accelerometers, are a cost-effective, portable alternative that may enhance opportunities for clinicians to measure pGRF during clinical and sport applications. PURPOSE: Examine various cut off frequencies associated with data collected using a wireless accelerometer compared to pGRF on force plates during four tasks. METHODS: Twenty-one healthy participants were recruited. Each were fit with an accelerometer secured in over the sacrum with a velcro belt. Tasks included a hang landing (HL), a countermovement jump (CMJ), running, and a single leg landing (SLL). Force plate data were collected at 2000 Hz and filtered at 50 Hz. Accelerometer data were collected at 1024 Hz and filtered systematically using a Butterworth filter with cutoff frequencies between 1-200 Hz. These data were compared to the force plate to find the most ideal frequency resulting in the lowest discrepancy between peaks of force. A repeated measures ANOVA examined differences in cutoff frequency between tasks (α = 0.05). Post hoc tests used the LSD approach. RESULTS: There were differences in cutoff frequencies among tasks (p = 0.002). Collectively, the CMJ and HL cutoff frequencies differed from the running and SLL (p < 0.05). Overall, the bilateral tasks required a higher cutoff frequency compared to unilateral tasks (Figure 1). CONCLUSIONS: Differences in cutoff frequencies between bilateral and unilateral tasks show that accelerometer data appear to be sensitive to the type of tasks. Researchers and clinicians should consider these cutoff frequency recommendations when selecting filtering parameters. Supported by: UWL Graduate Studies