Smart assistive wearable for gait rehabilitation using sensor fusion and real-time biofeedback (Note: If you’d like more options, I can provide a list.)

Project Abstract

This study presents the development and evaluation of a smart assistive wearable designed to enhance gait rehabilitation through sensor fusion and real-time biofeedback. The wearable integrates inertial measurement units (IMUs), pressure sensors embedded in insoles, and a compact gyroscopic module to capture multi-axial kinematics, ground reaction forces, and temporal-spatial gait parameters with high fidelity in real-world settings. A novel sensor fusion algorithm combines data from the IMUs and in-shoe pressure sensors to estimate joint angles, step length, cadence, step symmetry, and toe clearance, while onboard signal processing, aided by a lightweight neural network, enables low-latency detection of gait deviations indicative of asymmetry, compensatory strategies, or fatigue. Real-time biofeedback is delivered through multimodal cues, including haptic feedback via a wrist-worn actuator, auditory alerts, and a smartphone-based visual dashboard, allowing patients and clinicians to adjust rehabilitation tasks promptly and safely. The system supports adaptive therapy by incorporating patient-specific baselines and progression thresholds, enabling personalized progression of task difficulty, such as pace variation, obstacle negotiation, and dual-task walking. A modular design ensures compatibility with clinical and home environments, with wireless data transfer to a cloud-based analytics platform for longitudinal tracking and remote clinician oversight. The research employed a mixed-methods approach a pilot study with healthy volunteers to validate measurement accuracy against gold-standard optical motion capture and force plates, followed by a feasibility study with stroke and spinal cord injury patients to assess usability, comfort, and preliminary therapeutic efficacy. Quantitative outcomes demonstrated strong agreement with reference systems for step length (mean absolute error < 3 cm), gait speed (error < 0.05 m/s), and knee flexion angles (RMSE < 5 degrees) across varied walking speeds and terrains. In the patient cohort, participants showed statistically significant improvements in gait symmetry indices, swing-phase stability, and functional mobility scores after six weeks of home-based training complemented by clinician-guided sessions, compared to baseline and standard care controls. Qualitative feedback highlighted high acceptability of the wearable, with users reporting reduced cognitive load due to intuitive feedback modalities and a sense of increased confidence during ambulation. The study also examined safety, adherence, and fatigue, finding no adverse events and sustained engagement over the program duration. Data privacy and security considerations were addressed through end-to-end encryption, anonymization protocols, and opt-in data sharing controls aligned with applicable regulations. Limitations include sample size constraints, potential sensor drift during long-term use, and the need for further validation across diverse patient populations and locomotor impairments. The research contributes to the field by delivering an integrated, user-centered rehabilitation platform that leverages sensor fusion to deliver timely, personalized feedback, thereby accelerating motor relearning, promoting independence in mobility, and enabling scalable, home-based rehabilitation paradigms. Future work will explore advanced adaptive control strategies, transfer learning across patient groups, and integration with tele-rehabilitation ecosystems to broaden access and improve long-term outcomes.

Project Overview

What This Project Is About

---

In this project, we explore a wearable device that helps people improve how they walk after injury or illness. It uses sensors to collect movement data and smart software to give real-time feedback, guiding the user to move more normally and safely.

---

---

The Problem It Addresses

---

Many people struggle with gait problems after events like stroke or injury, leading to falls and reduced independence. Traditional rehab can be slow or boring. A wearable with simple feedback can make practice more effective and enjoyable, encouraging consistent training at home or in clinics.

---

---

Objectives of the Project

---

1. Explain how sensor data can reflect walking patterns in a way that is easy to understand.
2. Design a comfortable wearable that can be worn during daily activities.
3. Develop real-time feedback that helps users adjust their gait safely.
4. Test basic effectiveness through simple, user-friendly trials.

---

---

What You Will Do Step by Step

---

1) Review simple gait basics and choose non-technical sensors (like motion sensors). 2) Build a lightweight wearable prototype. 3) Collect movement data from volunteers while walking. 4) Create easy feedback signals (audio/visual) that guide steps. 5) Analyze improvements in walking style before and after practice. 6) Gather user feedback on comfort and usefulness. 7) Refine the device and feedback based on results. 8) Document the process and findings.

---

---

Expected Outcome

We expect a functional wearable that provides understandable feedback to help users improve gait, with preliminary evidence of improved walking patterns and high user acceptability for at-home use.