University of Twente Student Theses


Suitable metrics for upper limb movement smoothness during stroke recovery

Scheltinga, B.L. (2019) Suitable metrics for upper limb movement smoothness during stroke recovery.

[img] PDF
Abstract:Worldwide, about 10.3 million people have a first-ever stroke each year. It is estimated that only 5-20% of the patients show complete function recovery after 6 months post stroke. Movements of stroke patients are characterized by slowness, spatial and temporal discontinuity (non-smoothness) and abnormal patterns of muscle activation. Although there are multiple hypotheses, the exact mechanism why there is a lack of smoothness is unknown. Studying movement quality, for instance change in smoothness, during stroke recovery is vital to better understand the recovery process after stroke. Till so far, different smoothness metrics were used in research with stroke patients. It is seen that choices for the used metric are not always well motivated and even invalid metrics were introduced. In the first part of this thesis, all metrics that assess smoothness of reaching movements of stroke patients are obtained. It was seen that 31 different metrics have been used. By critical inspection of the mathematical expression of the metrics, it was seen that some metrics were mathematically invalid as they were dependent on movement duration or movement velocity. Next, velocity profiles of reaching movements were simulated while different parameters such as peak velocity or movement duration were changed. With these simulations, was seen that 9 out of the 31 metrics seemed suitable for assessing movement smoothness. Among these suitable metrics are four dimensionless squared jerk metrics, two metrics that work in the frequency domain of the velocity profile, a metric that compares the executed movement velocity profile to a standard velocity profile by means of correlation and a metric that counts the number of peaks in the velocity profile. As the metrics will be used with stroke patients, it is relevant to see if these metrics can capture changes in smoothness over time during the recovery of stroke patients. With clinical and kinematic data from 40 stroke subjects, measured multiple weeks after the stroke onset, two different linear mixed models using smoothness metrics were made. With these models, it was shown that the dimensionless jerk metrics performed slightly better in these models compared to the frequency spectrum based metrics. Another important finding in this master thesis is the difference between velocity profiles of reach-to-grasp movements and pointing movements. It was seen that the reach-to-grasp movements clearly have asymmetrical velocity profiles and the minimum jerk model does not apply for these movements.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:01 general works
Programme:Biomedical Engineering MSc (66226)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page