Optimal Spatial Sensor Design for Magnetic Tracking in a Myokinetic Control Interface.

Area of Science:

• Biomedical Engineering
• Robotics
• Sensor Technology

Background:

• Magnetic tracking is crucial for applications requiring object localization without a direct line-of-sight.
• Current challenges include optimizing sensor placement for maximum information and minimal computational load.
• Application to prosthetic hands involves tracking magnets in amputee forearm muscles for myokinetic control.

Purpose of the Study:

• To present the "Peaks method," a novel strategy for selecting an optimal sensor set for magnetic tracking.
• To minimize computational cost while maximizing information obtained from magnetic sensors.
• To enable intuitive control of prosthetic hands for amputees.

Main Methods:

• A 3D forearm model simulated proximal amputation with 11 implanted magnets.
• The Peaks method selected sensors from an initial grid by identifying peaks in magnetic flux density and its gradient.
• A calibration phase was included to customize sensor distribution to patient anatomy.

Main Results:

• The Peaks method selected 80 sensors, achieving localization accuracy below 0.22 mm (position) and 1.86° (orientation).
• Results were statistically comparable to a state-of-the-art alternative method, without requiring iterative solutions or prior knowledge.
• The calibration phase improved signal-to-noise ratio and adapted sensors to the patient's stump.

Conclusions:

• The Peaks method offers an efficient, generalizable solution for optimal sensor set design in magnetic tracking.
• It reduces localization computational cost by mirroring magnetic field shapes.
• This approach has broad applications in the biomedical field, particularly for prosthetic control.