Hybrid FEM and AI Framework Enhances Rehabilitation Monitoring Accuracy

Why It Matters

This approach moves beyond siloed biomechanical simulation and AI analysis by creating a closed-loop system. This allows for more precise validation of data acquisition and training of AI models, leading to a more adaptive and effective rehabilitation process.

Key Finding

A combined approach using detailed physical simulations and intelligent data analysis, supported by custom electronics, can effectively track and classify rehabilitation progress.

Key Findings

• The hybrid framework successfully integrates FEM simulation, AI classification, and embedded electronics.
• The AI algorithm demonstrated robust performance in classifying rehabilitation progress.
• The electronic system proved applicable in rehabilitation settings for real-time data acquisition and transmission.

Research Evidence

Aim: Can a hybrid framework combining FEM-based biomechanical simulation, AI trend classification, and embedded electronics improve the accuracy and adaptability of upper limb rehabilitation monitoring?

Method: Hybrid simulation and experimental validation

Procedure: A mechanical model of a robotic glove interacting with a deformable latex sphere was developed using FEM. This simulation data was used to validate a custom electronic system for signal acquisition and wireless data transmission. The acquired biomechanical features were then used to train a supervised AI algorithm for classifying rehabilitation progress.

Context: Rehabilitation engineering, biomechanics, medical device development

Design Principle

Synergistic integration of simulation, AI, and embedded systems enhances the intelligence and effectiveness of monitoring and rehabilitation tools.

How to Apply

When designing assistive or rehabilitative devices, use simulation to create realistic scenarios for training AI algorithms that interpret user biomechanics.

Limitations

The study used a simplified latex sphere model; real human tissue complexity may differ. The AI model's generalizability to diverse patient populations and conditions requires further investigation.

Student Guide (IB Design Technology)

Simple Explanation: By combining computer simulations of how things move with smart computer programs (AI) and special sensors, we can build better tools to help people recover from injuries, like making sure they are doing their exercises correctly.

Why This Matters: This research shows how combining different advanced technologies can lead to more effective and personalized healthcare solutions, which is a key area for design innovation.

Critical Thinking: How does the choice of material model (e.g., Mooney-Rivlin) in FEM affect the accuracy of the AI's classification of rehabilitation progress?

IA-Ready Paragraph: The integration of finite element modelling (FEM) with artificial intelligence (AI) and embedded electronics, as demonstrated in studies like Laganà et al. (2025), offers a powerful paradigm for developing advanced monitoring systems in rehabilitation. By using high-fidelity simulations to generate realistic biomechanical data, designers can train AI algorithms to accurately classify patient progress, leading to more adaptive and effective therapeutic interventions.

Project Tips

• When developing a rehabilitation device, consider how simulation data can be used to train AI for performance monitoring.
• Explore the use of embedded systems for real-time data collection to provide immediate feedback.

How to Use in IA

• This research can inform the development of a novel monitoring system for a rehabilitation device, justifying the use of simulation and AI for data analysis and feedback.

Examiner Tips

• Evaluate the justification for using a hybrid modelling approach over simpler methods.
• Assess the clarity of the connection between simulation outputs and AI training data.

Independent Variable: ["Biomechanical features derived from FEM simulation and sensor data"]

Dependent Variable: ["Classification of rehabilitation progress (improving/worsening)"]

Controlled Variables: ["Material properties (Mooney-Rivlin parameters)","Contact conditions","Data acquisition parameters","AI algorithm architecture"]

Strengths

• Novel integration of FEM, AI, and embedded electronics.
• Validation of simulation data against real-world signal acquisition.

Critical Questions

• What are the computational costs associated with high-fidelity FEM simulations for real-time applications?
• How can the AI model be made robust to variations in sensor noise and individual patient biomechanics?

Extended Essay Application

• Investigate the use of simplified FEM models or machine learning surrogate models to reduce computational load for real-time AI training in rehabilitation devices.
• Explore different AI architectures for improved classification accuracy and generalization across diverse user groups.

Source

FEM-Based Modelling and AI-Enhanced Monitoring System for Upper Limb Rehabilitation · Electronics · 2025 · 10.3390/electronics14112268