IMU-Based Knee Angle Estimation using an Extended Kalman Filter.

Area of Science:

• Biomedical Engineering
• Rehabilitation Robotics
• Wearable Sensor Technology

Background:

• Robotic rehabilitation for hemiparetic patients often requires monitoring the healthy leg using ambulatory systems like Inertial Measurement Units (IMUs).
• IMUs present challenges including numerical drift, magnetic field interference, and critical sensor alignment issues.
• Accurate knee angle estimation is crucial for effective gait analysis and rehabilitation feedback.

Purpose of the Study:

• To develop and validate a novel algorithm for precise knee angle estimation in ambulatory settings.
• To address the limitations of traditional IMU-based sensing in robotic rehabilitation.
• To provide a reliable method for monitoring lower limb kinematics during rehabilitation exercises.

Main Methods:

• An algorithm based on a recursive paradigm to estimate the rotational axis was developed.
• Geometrical constraints were integrated with an Extended Kalman Filter for knee angle estimation.
• The algorithm was tested on five healthy subjects to assess its performance.

Main Results:

• The algorithm successfully reconstructed the knee angle waveform.
• The average root-mean-square (rms) error was less than 2 degrees.
• The method demonstrated robustness against common IMU sensor drawbacks.

Conclusions:

• The proposed algorithm offers an accurate and reliable solution for knee angle estimation.
• This technology can enhance the effectiveness of robotic-based rehabilitation for hemiparetic individuals.
• The algorithm provides a foundation for improved ambulatory monitoring in clinical settings.