Assessing Personalized FES Patterns for Gait Rehabilitation in a Clinical Setting

Abstract

Functional electrical stimulation (FES) for motor and neurological rehabilitation produces patient-dependent functional responses influenced by the individual's pathology and clinical profile. Given this inherent variability in patient responses, it is essential to personalize FES gait rehabilitation parameters to enhance the rehabilitation outcomes effectively. To optimize these parameters for gait neurorehabilitation, appropriate metrics are needed. In this preliminary case study, we assess if selected metrics can comprehensively evaluate the effectiveness of FES gait patterns with different degrees of personalization for specific post-stroke patients in the context of a clinical rehabilitation routine. These metrics do not only reflect the immediate orthotic effects, but also gait parameters that indicate rehabilitation progress and improved treatment outcomes. We analyzed data collected during rehabilitation of three post-stroke hemiplegic individuals with NeuroSkin, a multichannel textile FES system with embedded movement sensors and dry electrodes. We evaluated different stimulation patterns using metrics related to kinematic, spatio-temporal, postural, and asymmetry aspects of gait. Some metrics identified modest overall improvements between default and personalized stimulation patterns: walking speed (+0.63 cm/s) and stride length (+0.41 cm). However, patient-dependent analysis revealed substantial improvements which differed between patients across walking speed (+3.3 cm/s), stride length (+4.7 cm) and metrics reflecting asymmetry and postural control, as well as kinematic trajectories. We argue that these metrics should be integrated in FES-aided gait assessments. However, it remained difficult to identify which metrics should be primarily considered when optimizing an FES strategy, given heterogeneous improvements and limitations of the current framework.