The FOSTER protocol: Mapping postural control and functional recovery after stroke using clinical and sensor-based assessments
 
The FOSTER protocol: Mapping postural control and functional recovery after stroke using clinical and sensor-based assessments ■
 
Charlotte Moeyersons, Britta Hanssen, Elissa Embrechts, Mahyar Firouzi, Sylvie De Raedt, , Bernard Dan, Bart Jansen, Marc Degelaen, Eva Swinnen
 
Abstract ■

Introduction: Stroke is a leading cause of long-term disability, with recovery trajectories that are highly heterogeneous and difficult to predict at the individual level [1]. Standardized clinical assessments primarily capture task accomplishment under controlled conditions, but provide limited insight into movement quality, compensatory strategies, and the postural control mechanisms underlying functional performance. Recent consensus recommendations emphasize the need for sensitive, functionally relevant measures of balance and mobility [2,3]. Integrating clinical assessments and sensor-based measurements during functional tasks enables a more detailed characterization of recovery. Research Question: This study aims to (1) characterize longitudinal trajectories of postural control, functional mobility, and activity after stroke using combined clinical and sensor-based assessments, and (2) examine the relationships between sensor-derived measures of movement quality and conventional clinical outcomes over time. Methods: The FOSTER study is a prospective, single-centre longitudinal cohort study conducted in a specialized rehabilitation hospital. Adults within 8 weeks post-stroke are assessed at up to five time points (baseline (time of admission), 5, 8, 12, and 24 weeks post-stroke) (Fig. 1A). The protocol combines standardized clinical assessments, patient-reported outcomes, and synchronised sensor-based measurements during functional tasks, including seated reaching, standing reaching, and gait initiation with an interactive exergame (Fig. 1B). The clinical assessment battery includes measures of lower limb motor impairment (Fugl-Meyer Assessment – Lower Extremity, Motricity Index), trunk control (Trunk Impairment Scale), balance (Mini-BESTest and/or Berg Balance Scale), gait performance (10-Meter Walk Test, Functional Ambulation Category) and global disability (modified Rankin Scale). Patient-reported outcomes include the Hospital Anxiety and Depression Scale and the Short Form-36. Sensor modalities include surface electromyography, inertial measurement units, force plates, and markerless motion capture, providing synchronized kinetic, kinematic, and neuromuscular data. Real-world physical activity is monitored using wearable accelerometry (Activ8) over 72 hours at each time point. Longitudinal trajectories and associations between clinical and sensor-derived measures will be analysed using linear mixed-effects modelling (Fig. 1C). Figure 1. Image 69 Results: Data collection is ongoing and spans January 2025 through December 2027, aiming to enrol a total of 100 patients. To date, 48 participants have been enrolled, accounting for 154 completed assessments, with 15 participants currently in active follow-up. Planned analyses will describe longitudinal changes in postural control, functional mobility, and activity, and explore associations between clinical outcomes and sensor-derived measures of movement quality. Discussion: This study provides a standardized and clinically embedded framework to investigate recovery and underlying mechanisms of postural control and functional activity after stroke. By integrating multimodal sensor data with established clinical measures across multiple time points, this innovative protocol enables a more comprehensive understanding of recovery processes and inter-individual variability. Within the wider EU-funded TARGET project [4], this will contribute to development of data-driven, individualized models, including {"}virtual twin{"}, to support personalized rehabilitation and improve outcome prediction. References: [1] Todhunter-Brown A, et al. Cochrane Database Syst Rev. 2025. [2] Kwakkel G, et al. Eur Stroke J. 2023. [3] Van Criekinge T, et al. Neurorehabil Neural Repair. 2024. [4] Ortega-Martorell S, et al. Eur Heart J 2025.