Discrepancies between plantar pressure devices: Evaluating cross-system reliability for biomechanics, clinical use and predictive modelling

Abstract

Plantar pressure measurement systems are widely used to assess foot function and gait, yet discrepancies in sensor design, measurement protocols, and population characteristics can undermine data comparability. This study investigated three platform‑based and two in‑shoe systems to evaluate key parameters such as the contact area, maximum force, force‑time integral, peak pressure, pressure‑time integral, maximum mean pressure and contact time. Fifteen healthy adults walked at a self‑selected pace, providing a total of 360 footprints from the platforms (barefoot) and 1200 footprints from the in‑shoe devices (shod). Each footprint was then divided into hindfoot, midfoot, and forefoot regions. A two‑way repeated‑measures ANOVA (systems × regions) revealed that mean values (MV) and coefficients of variation (CV) frequently differed among devices, indicating limited cross‑system comparability. Moreover, intraclass correlation coefficients for peak pressure ranged between poor (<0.5) and, on rare occasions, moderate (0.5–0.75), further confirming substantial variability. These discrepancies highlight the importance of standardising calibration, data extraction, and analysis protocols, as even devices based on similar resistive or capacitive technologies can produce dissimilar outcomes. Environmental factors such as footwear selection and lab‑based “targeting” errors also contribute to inconsistencies. These challenges are especially relevant as emerging technologies integrate high-resolution wearable sensors with artificial intelligence to support real-time clinical decision-making, disease prediction and personalised interventions. Establishing uniform reporting and validation standards will be essential to ensure robustness and comparability in both traditional biomechanical studies and future AI-driven applications.