Electromyography (EMG)

• EMG stands for Electromyography
• Electromyography studies the condition of muscles and nerves using electrical signals from the muscles.
• The primary EMG component is Motor unit action potentials (MUAPs).
• MUAP is a collection of action potentials released from muscle fibers.

Recording EMG

• Detects the electrical activity and transform ionic to electric current
• Easy to detect but hard to analyse because of noise.
• There are two types: surface EMG and Intramuscular EMG
• The sensor should be located in the middle of the muscle belly and in orientation of the muscle fibre.
• Has amplitude and frequency
• Characteristics are expressed in the time and frequency domains.

Types of Noise in EMG recordings

• Internal noise includes fiber type, blood flow, subcutaneous fat, and conduction velocity
• External noise includes skin impedance, muscle crosstalk, 50 Hz noise, and sensor movement

EMG Analysis example

• Can be performed for Isometric contraction for 120 seconds
• Isometric contraction analysis included the frequency domain of the last and first seconds
• Analysis can be performed for Isotonic contraction as well.
• Isotonic contraction is where the muscle length changes in order to produce movement
• Can be performed for Isotonic contraction for Multi-channel using multiple channel recordings simultaneously
• Spectrum analysis for each of them is available

EMG Signal Processing Operations

• Band pass filter / Notch filter
• Rectification
• Root mean square envelope
• Normalisation

EMG for Prosthetics Control

On-off Myoelectric Control

• Requires a threshold of EMG to control the hand
• Operates the hand at a constant speed unrelated to the level of contraction

Proportional Myoelectric Control

• Applies voltage to the motor proportional to the contraction level or intensity of EMG signals
• Used for gross movement

Direct Myoelectric Control

• Independent EMG channels are similar to proportional control for individual and direct movements

Finite State Machine Control

• The postures of the hands are predefined as states
• The transition among states is also predefined from the inputs

Terminal Devices based on DoFs/Motors

• Can be categorised according to DoFs/Motors: single DoF vs Multi-DoFs
• Single-DoF uses one motor and only opens/closes, is durable, and generates the most pinch force
• Multiarticulate uses more than one motor for more grasping patterns, is less durable, and has less pinching force

Terminal Devices based on Control

• Can be categorised according to control: proportional vs digital
• Proportional control angle of motor is directly proportional with the exerted force/EMG and is easier to learn and adjust the force
• Digital = ON/OFF control
• Consists of programmed grasping functions, is not intuitive, some proportional controls can be included