Task independent identification of sensor location on upper limb from orientation data

Area of Science:

• Biomedical Engineering
• Neuroscience
• Rehabilitation Technology

Background:

• Tremor is the most common movement disorder, necessitating effective long-term monitoring and suppression strategies.
• Current neuroprosthetic applications often have fixed sensor locations, limiting adaptability.
• Understanding sensor placement is crucial for accurate movement disorder analysis.

Purpose of the Study:

• To develop and validate a novel method for sensor placement identification in upper limb neuroprostheses.
• To enhance the effectiveness of tremor suppression and facilitate long-term tremor monitoring.
• To simplify classification algorithms for neuroprosthetic applications.

Main Methods:

• A novel sensor placement identification method was developed based on relative sensor location.
• The method was tested on an upper limb neuroprosthesis for tremor suppression in patients with essential tremor or Parkinson's disease.
• Seventeen different tasks were performed to evaluate the method's robustness.

Main Results:

• The proposed method achieved an average accuracy of 98.30% for tremor suppression across various tasks.
• Ten key features were identified for optimal sensor placement and classification.
• The system demonstrated high accuracy irrespective of the specific task performed.

Conclusions:

• The novel sensor placement identification method is effective for tremor suppression and monitoring using neuroprostheses.
• This approach simplifies the use of wearable sensors for non-trained personnel, promoting user-friendliness.
• The findings represent a significant advancement in developing context-aware neuroprosthetic devices.