fNIRS-Based Upper Limb Motion Intention Recognition Using an Artificial Neural Network for Transhumeral Amputees

Area of Science:

• Neuroscience and Biomedical Engineering
• Brain-Computer Interfaces (BCI)
• Prosthetics and Rehabilitation

Background:

• Prosthetic arms aid daily activities for amputees, but current brain-machine interfaces have limitations in motion prediction and rehabilitation for transhumeral amputations.
• Enhanced control and accuracy for upper limb prostheses are crucial for improving amputees' quality of life.

Purpose of the Study:

• To propose and evaluate a functional near-infrared spectroscopy (fNIRS)-based approach for recognizing human intention across six upper limb motions.
• To investigate the potential of fNIRS signals for real-time control of transhumeral prostheses.

Main Methods:

• Acquired fNIRS signals from the motor cortex of 15 healthy subjects and 3 transhumeral amputees during six specific upper limb movements.
• Filtered fNIRS data using a finite impulse response (FIR) filter and extracted features including signal mean, peak, and minimum values.
• Utilized an artificial neural network (ANN) for classifying the intended arm actions.

Main Results:

• The fNIRS-based approach achieved a 78% accuracy in classifying six distinct upper limb arm actions.
• The classification accuracy for prosthetic arm control using fNIRS signals is a novel finding in this research area.

Conclusions:

• The proposed fNIRS method demonstrates significant potential for detecting human intention for upper limb movements.
• These promising results suggest the feasibility of applying fNIRS for the real-time control of transhumeral prosthetic arms, advancing rehabilitation technology.