Estimation of physiological tremor from accelerometers for real-time applications

Area of Science:

• Robotics
• Biomedical Engineering
• Signal Processing

Background:

• Accurate filtering of physiological tremor is critical for robotic-assisted surgery.
• Existing tremor filtering methods often assume a single dominant frequency, which is insufficient for real-world physiological tremor.
• Time-frequency analysis reveals physiological tremor comprises multiple dominant frequencies.

Purpose of the Study:

• To develop robust, single-stage algorithms for accurate physiological tremor filtering using accelerometers for real-time robotic surgery applications.
• To address the limitations of existing methods by accounting for the multi-frequency nature of tremor.
• To formulate a procedure for separating intended motion from tremor.

Main Methods:

• Review of existing tremor filtering algorithms.
• Development of two novel, improved algorithms for tremor filtering.
• Comparative analysis of new and existing algorithms using accelerometer data from microsurgeons and novice subjects.
• Time-frequency analysis of physiological tremor data.

Main Results:

• The developed algorithms demonstrated superior performance in tremor estimation compared to existing methods.
• The new algorithms effectively filter physiological tremor containing multiple dominant frequencies.
• A method for distinguishing intended motion/drift from tremor was successfully formulated.

Conclusions:

• The proposed algorithms offer a more accurate and robust solution for real-time physiological tremor filtering in robotic surgery.
• Accounting for multiple tremor frequencies is essential for effective tremor suppression.
• These advancements can improve the safety and efficacy of robot-assisted surgical interventions.