Machine-Learning-Based Video Analysis of Grasping Behavior During Recovery from Cervical Spinal Cord Injury

Behav Brain Res, 2023 · DOI: 10.1016/j.bbr.2022.114150 · Published: April 12, 2023

Simple Explanation

This study uses machine learning to analyze videos of monkeys recovering from spinal cord injuries. The goal was to sensitively quantify kinematic aspects of grasping behavioral deficits. The researchers tracked the movements of the monkeys' fingers as they grasped sugar pellets. They found that even when the monkeys seemed to recover their ability to grab the pellets, their finger movements were still not quite normal. This suggests that even after a spinal cord injury, traditional methods for assessing recovery may not be sensitive enough to detect subtle but persistent deficits in fine motor control.

Study Duration

14 Weeks

Participants

3 adult male squirrel monkeys

Evidence Level

Not specified

Key Findings

1
Cervical dorsal column lesion (DCL) produces a change in digit orientation from lateral to vertical.
2
DCL causes an increased variance in digit movements, especially in the vertical plane.
3
After DCL, a longer time is required to complete the grasping task.

Research Summary

The study used machine-learning-based video analysis to quantify fine-grained finger behavior during a sugar pellet grasping task after a unilateral cervical DCL. The analysis identified a change in digit orientation, increased variance in digit movement, and longer time to complete the task after DCL, which persisted through 14 weeks. Traditional end-point analyses showed recovery of successful retrievals by 14 weeks, highlighting the need for combined global and fine-scale kinematic behavioral analysis.

Practical Implications

Improved Assessment of Recovery

Combining traditional end-point measures with machine-learning-based kinematic analysis provides a more comprehensive assessment of grasping behavior recovery after spinal cord injury.

Targeted Therapeutic Interventions

The sensitive identification of fine-scale kinematic deficits can aid in the development of targeted therapeutic interventions to improve hand sensorimotor behavior.

Clinical Translation

The DeepLabCut-derived kinematic measures can be applied to human grasping behavior during the Pegboard Dexterity Test, allowing for direct comparison of animal and human studies.

Study Limitations

1
Relatively small sample size
2
Reliance on cooperative squirrel monkey behavior
3
The difficulty of the behavioral task used in the current study - Kluver Board test

Your Feedback

Was this summary helpful?

Back to Spinal Cord Injury