Skip to content

Latest commit

 

History

History
45 lines (31 loc) · 3.05 KB

README.md

File metadata and controls

45 lines (31 loc) · 3.05 KB

Differentiable Biomechanics for Markerless Motion Capture in Upper Limb Stroke Rehabilitation: A Comparison with Optical Motion Capture

Tim Unger*1, 4, Arash Sal Moslehian*2, J.D. Peiffer*3, Johann Ullrich1, 4, Roger Gassert4, Olivier Lambercy4, R. James Cotton†5 , Chris Awai Easthope†1

1Data Analytics & Rehabilitation Technology (DART Lab), Lake Lucerne Institute, Vitznau, Switzerland
2Neuro-X Institute, EPFL, Lausanne, Switzerland
3Dept of Biomedical Engineering, Northwestern University, Chicago, IL
4Rehabilitation Engineering Laboratory (RELab), ETH, Zurich, Switzerland
5Shirley Ryan AbilityLab, Dept of PM&R, Northwestern University, Chicago, IL
*co-first author
†co-senior author

This repository includes the analysis code of our ICORR 2025 Paper.

Abstract

Marker-based Optical Motion Capture (OMC) paired with biomechanical modeling is currently considered the most precise and accurate method for measuring human movement kinematics. Combining differentiable biomechanical modeling with Markerless Motion Capture (MMC) offers a promising approach to motion capture in clinical settings, requiring only minimal equipment, such as synchronized webcams, and minimal effort for data collection.

This study compares key kinematic outcomes from biomechanically modeled MMC and OMC data in 15 stroke patients performing the drinking task, a functional task recommended for assessing upper limb movement quality.

We observed a high level of agreement in kinematic trajectories between MMC and OMC, as indicated by high correlations (median $r > 0.95$ for the majority of kinematic trajectories) and median $\text{RMSE}$ values ranging from $2^\circ$– $5^\circ$ for joint angles, $0.04 , \text{m/s}$ for end-effector velocity, and $6 , \text{mm}$ for trunk displacement. Trial-to-trial biases between OMC and MMC were consistent within participant sessions, with interquartile ranges of bias around $1-3^\circ$ for joint angles, $0.01m/s$ in end-effector velocity, and approximately $3 , \text{mm}$ for trunk displacement.

Our findings indicate that our MMC for arm tracking is approaching the accuracy of marker-based methods, supporting its potential for use in clinical settings. MMC could provide valuable insights into movement rehabilitation in stroke patients, potentially enhancing the effectiveness of rehabilitation strategies.

The Code

Will be hosted here after the publication.

Dataset

We are actively working on getting the rights to publishing the anonymized patient dataset for this work. Once done, the data will be linked here.

Citation

@inproceedings{tba,
  title={{tba}},
  author={tba},
  booktitle={tba},
  year={2025}
}