Simulating acceleration from stereophotogrammetry for medical device design.

Area of Science:

• Biomechanics
• Medical Device Design
• Sensor Technology

Background:

• Designing medical devices with accelerometers requires optimizing sensor placement, quantity, and location for performance and cost.
• Stereophotogrammetry offers a non-portable method to capture motion data for simulating accelerometer measurements.

Purpose of the Study:

• To investigate the applicability of simulated acceleration signals for medical device design.
• To evaluate the error associated with simulated accelerometer signals under various conditions.
• To demonstrate the use of simulated signals in system design applications.

Main Methods:

• Utilized stereophotogrammetry to measure the dynamics of a reference coordinate frame.
• Simulated "virtual" accelerometers at different body locations using captured dynamics.
• Compared simulated signals with directly measured accelerometer data to assess error.
• Applied simulated signals in a medical device system design example.

Main Results:

• The best-case error between a virtual and measured signal was 0.028 m/s².
• Errors increased significantly when using poorly estimated centripetal and tangential acceleration terms.
• Limb non-rigidity can dramatically increase error, but modeling can mitigate this.

Conclusions:

• Simulated acceleration signals offer benefits for medical device design when accurate body segment models are employed.
• Careful modeling is crucial for reducing errors caused by limb non-rigidity and acceleration estimation.
• Virtual accelerometer simulation is a valuable tool for optimizing accelerometer-based medical device development.