Skip to main content

Table 3 Benchmarking indicators for motor abilities

From: A unified scheme for the benchmarking of upper limb functions in neurological disorders

Motor Ability

Performance Indicator

Domain

Mandatory/Recommended

Description

Relevant reference

Kin

EMG

Accuracy

Trajectory absolute error

x

 

R

Mean or maximal distance between ideal and actual trajectory between movement onset and end

[57]

End-point error

x

 

R

Mean or maximal Euclidean distance between actual and target position at or after movement end

[58]

Variable error

x

 

R

Standard deviation of the end-point error across multiple repetitions of the movement or task

[59]

Area Index

x

 

R

Area between the desired straight line and the path actually performed

[60]

Efficacy

Success rate

x

 

R

Number of accomplished objectives (e.g., movements performed, tasks completed) divided by the total number of attempts

[61]

Number of movements stops

x

 

M

Number of times that the velocity curve dropped below a percentage of peak velocity after movement onset

[62]

Efficiency

Movement time

x

 

R

Time from the onset to the end of a task or movement

[63]

Path traveled

x

 

R

Path length covered between onset and end of a movement or task

[64]

Path length ratio

x

 

M

Ratio between the path traveled and the shortest possible distance between movement onset and end

[65]

Trunk compensation

x

 

R

Ratio between trunk displacement and hand displacement in the sagittal plane

[66]

Waveform length

 

x

R

Cumulative length of the waveform of the EMG signal from the ith-muscle over the time segment

[67]

Average amplitude change

 

x

R

Mean of the cumulative length of the waveform of the EMG signal from the ith-muscle over the time segment

[55]

Difference absolute standard deviation value

 

x

R

Standard deviation of the cumulative length of the waveform of the EMG signal from the ith-muscle over the time segment

[55]

Intra-limb coordination

Joint angle correlation

x

 

R

Correlation between shoulder flexion–extension and elbow flexion–extension joint time angles profiles

[68]

Elbow peak velocity

x

 

R

Highest value of the elbow flexion/extension joint velocity profile during movement

[43]

Time to peak elbow extension angle

x

 

R

Time to reach peak extension angle for the elbow joint, relative to the duration of the movement

[69]

Muscular synergies

 

x

R

Linear decomposition algorithm (e.g., principal components analysis, factor analysis, independent component analysis, and non-negative matrix factorization) to extract spatiotemporal, temporal, and spatial features from EMG signal of the muscles mainly involved in the task

[70]

Co-contraction index

 

x

R

Percentage of overlapping activity of EMG linear envelopes between the agonist and the antagonist muscle involved in the task

[71]

 

Intermuscular coherence

 

x

R

Square of the cross-spectra normalized with auto-spectra derived from the EMG signal from the agonist and the antagonist muscles involved in the task

[72]

Movement amplitude

Joint range of motion

x

 

M

Range of the anatomical joints angles between movement onset and end

[73]

Maximum reached distance

x

 

R

Maximum distance reached from the starting position

[74]

Trunk displacement

x

 

R

Euclidean distance covered by the trunk between movement onset and end

[75]

Normalized reaching area

x

 

R

Maximally reached or reachable position during a movement or task divided by the length of the user’s arm

[76]

Muscular effort

Integrated EMG

 

x

R

Summation of rectified EMG signal amplitude obtained from the ith-muscle

[55]

Root Mean Square

 

x

R

Square root of the mean square of the EMG signal amplitude obtained from the ith-muscle

[55]

Activation level

 

x

M

Average of the absolute value of the EMG signal amplitude obtained from the ith-muscle in a segment

[77]

Variance of EMG

 

x

R

Average of squared EMG signal amplitude obtained from the ith-muscle

[67]

Mean absolute value slope

 

x

R

Differences between mean absolute values of the EMG signal amplitude obtained from the ith-muscle of the adjacent segments

[55]

Planning predictability

Time to peak velocity

x

 

R

Time to reach peak velocity relative to the duration of the movement

[78]

Reaction time/Response latency

x

 

M

Time between the “Go” cue (as indicated by visual/acoustic feedback or any other channel) of a movement and the actual onset of the movement (e.g., 10% of peak velocity)

[62]

Muscle onset

 

x

R

Time between the “Go” cue and the onset (e.g., detected by the Teager–Kaiser energy operator) of the EMG signal amplitude obtained from the ith-muscle

[79]

Initial movement direction error

x

 

R

Distance between ideal and actual trajectory at an initial time point right after movement onset (e.g., 10% of peak velocity)

[62]

Aiming angle

x

 

R

Angular difference between target direction and direction of travel calculated from starting point up to peak speed point

[80]

Power

Mean frequency

 

x

R

Sum of the product of the power spectrum of the EMG signal from the ith-muscle and the frequency divided by the total sum of the spectrum intensity

[56]

Median frequency

 

x

R

Frequency at which the power spectrum of the EMG signal from the ith-muscle is divided into two regions with equal amplitude

[56]

Mean power

 

x

R

Average power of the power spectrum of the EMG signal from the ith-muscle

[55]

Power spectral density

 

x

R

Amount of power per frequency interval of the power spectrum of the EMG signal from the ith-muscle

[56]

Frequency ratio

 

x

R

Ratio between the low-frequency components and the high-frequency components of the power spectrum of the EMG signal from the ith-muscle

[55]

Power spectrum ratio

 

x

R

Ratio between the energy which is nearby the maximum value of the power spectrum of the EMG signal from the ith-muscle, and the energy which is the whole energy of the power spectrum

[55]

Smoothness

Number of velocity peaks

x

 

M

Number of peaks (i.e., maxima above a certain threshold) in the velocity profile between movement onset and end

[81]

Speed correlation to idealized profile

x

 

R

Correlation between actual speed profile and idealized normal velocity profile (e.g., straight line)

[82]

Movement Arrest Period Ratio

x

 

R

Proportion of time that movement speed exceeds a given percentage of peak speed

[83]

Peak Speed Ratio

x

 

R

Mean speed divided by the peak speed

[83]

Normalized dimensionless jerk

x

 

R

Time-integral of the squared jerk (i.e., third time-derivative of position) between movement onset and end normalized with respect to movement duration to the power of five and movement length to the power of two

[84]

Spectral arc length

x

 

R

Length of the spectral trajectory (i.e., in the frequency domain) of the velocity profile between movement onset and end

[85]

Mean acceleration

x

 

R

Mean value of the acceleration profile between movement onset and end

[86]

EMG Zero Crossing

 

x

R

Number of times that amplitude values of the EMG signal from the ith-muscle crosses zero amplitude level

[55]

Slope sign change

 

x

R

Number of times that slope of the EMG signal from the ith-muscle changes sign

[55]

Speed

Peak velocity

x

 

R

Maximal value of the velocity profile between movement onset and end

[87]

Mean velocity

x

 

M

Mean value of the velocity profile between movement onset and end

[83]

Mean velocity variability

x

 

R

Difference between the velocity profile of the participant’s reaching trajectory and the ideal velocity profile for each movement

[80]

  1. If not specified, kinematic outcomes must be computed from a distal joint or the robot end-effector. Electromyography outcomes must be computed for every muscle recorded if not specified. Relevant reference specifies the first work of the literature that, to our knowledge, used that specific outcome