Biomechanical analysis and rehabilitation protocol optimization for post-stroke upper limb motor recovery using wearable inertial sensors

Project Abstract

This study presents a multidisciplinary approach to enhancing post-stroke upper limb motor recovery through a biomechanically informed rehabilitation protocol augmented by wearable inertial sensors. The primary objective is to quantify kinematic and kinetic impairments in the paretic upper limb, translate these metrics into individualized therapy targets, and optimize rehabilitation regimens to maximize gait-like symmetry, range of motion, and functional task performance within a home- and clinic-based setting. A mixed-methods design integrates a cross-sectional biomechanical assessment, a longitudinal intervention, and a qualitative evaluation of user experience and adherence. We recruit adult stroke survivors (12–24 weeks post-onset) with varying degrees of unilateral upper limb impairment, alongside a healthy control cohort to establish normative baselines. Wearable inertial measurement units (IMUs) deployed on the forearm, wrist, and upper arm capture high-fidelity data on joint angles, velocity, acceleration, jerk, and coordination patterns during standardized tasks (reaching, grasping, multi-step object manipulation) as well as functional activities of daily living. A custom data pipeline converts IMU data into clinically interpretable metrics such as joint torque proxies, smoothness indices, inter-joint coordination measures, and muscle activation estimates derived from motion proxies. These metrics inform a decision-support framework that tailors rehabilitation protocols by modulating task difficulty, feedback modality, and assist-as-needed support from a robotic exoskeleton or therapeutic assist devices. The intervention comprises 8–12 weeks of thrice-weekly sessions, combining task-specific functional practice, proprioceptive and motor imagery training, and repetitive, progressive resistance exercises prescribed via a tablet interface. We hypothesize that IMU-guided, individualized rehabilitation will produce greater gains in motor impairment (as measured by the Fugl-Meyer Assessment for the upper extremity), functional activity (Wolf Motor Function Test), and quality-of-life indices compared to conventional therapy alone. Secondary outcomes include changes in kinematic synergy, movement smoothness (log dimensionless jerk), and time-to-task completion, alongside adoption metrics such as user engagement, perceived ease of use, and adherence. A nested qualitative component with semistructured interviews will explore perceived barriers and facilitators to home-based wearables, perceived usefulness of feedback, and the acceptability of the tailored protocol across different severities. Statistical analyses will employ mixed-effects models to account for repeated measures and individual variability, with subgroup analyses by severity, age, and lesion laterality. Ethical considerations cover data privacy, device safety, and informed consent. Anticipated contributions include (1) a validated, scalable workflow for biomechanically grounded rehabilitation using readily available wearables, (2) evidence on the efficacy of personalized, feedback-rich protocols in accelerating motor recovery, and (3) a translational framework for clinicians to integrate IMU-derived metrics into routine practice. Potential challenges include sensor calibration drift, data synchronization, and ensuring user adherence in home settings, which will be mitigated through robust preprocessing, offline validation, and user-centered interface design. The study aims to bridge laboratory-based biomechanics and practical, real-world rehabilitation to improve functional independence for individuals recovering from stroke.

Project Overview

What This Project Is About

---

A plain-language overview of how looking at arm movements after a stroke can help recovery. The project uses wearable sensors to track how the upper limb moves and then tests ways to improve rehabilitation protocols based on that data.

---

---

The Problem It Addresses

---

Many stroke survivors have lasting difficulty with arm movement, and current rehab methods may not be tailored to each person. This project aims to fill gaps by measuring movement in everyday tasks and using those measurements to customize therapy plans.

---

---

Objectives of the Project

---

1. Understand how the arm moves after stroke using simple sensor data.
2. Identify patterns that indicate better or poorer recovery progress.
3. Develop a basic, practical rehabilitation protocol that can be tried in clinics.
4. Assess whether the protocol improves movement quality in a short trial.

---

---

What You Will Do Step by Step

---

1. Review basic stroke rehabilitation ideas and wearable sensor basics.
2. Collect movement data from participants performing everyday tasks with a wearable sensor on the affected arm.
3. Analyze data to find meaningful movement patterns and improvements over time.
4. Design simple rehab activities aligned with data insights.
5. Run a small pilot to test the new protocol's feasibility.
6. Summarize findings and suggest future improvements or larger studies.

---

---

Expected Outcome

Anticipated results include a clearer link between sensor-measured movement and recovery progress, plus a ready-to-use, data-informed rehabilitation protocol that clinicians can adapt to individual patients.