Validation of instrumental indices for the upper limb function assessment in neurological patients

differences between the SP estimates and those obtained fromeach testedmethodwere calculated. Themethods’ sensitivity and specificity were estimated by counting the number of extra and missed GEs with respect to the true GEs. The robustness of the methods with respect to the IMU position was assessed by calculating the mean percent difference (% pos) between obtainedwith the IMU positioned in the recommended location and that obtained in the other locations.