Parkinson’s disease screening using a fusion of gait point cloud and silhouette features
2025

Screening for Parkinson's Disease Using Gait Features

Sample size: 294 publication Evidence: moderate

Author Information

Author(s): Connie Tee, Aderinola Timilehin B., Ong Jia You, Ong Thian Song, Goh Michael Kah Ong, Erfianto Bayu, Purnama Bedy, Lim Ming De, Saedon Nor Izzati

Primary Institution: Multimedia University, Malacca, Malaysia

Hypothesis

Can a fusion of gait point cloud and silhouette features improve the screening for Parkinson's Disease?

Conclusion

The study demonstrates that combining model-free and model-based gait features significantly enhances the screening accuracy for Parkinson's Disease.

Supporting Evidence

  • The fusion of gait features achieved an AUC of up to 0.87.
  • F1-scores reached up to 0.82 using Logistic Regression.
  • The study highlights the potential of AI in early detection of Parkinson's Disease.

Takeaway

This study shows that we can use videos of people walking to help find out if they might have Parkinson's Disease, which can help them get treatment sooner.

Methodology

The study used videos to capture gait, extracted keypoint coordinates, and performed binary classification of gait as normal or Parkinsonian using a fusion of gait point cloud and silhouette features.

Potential Biases

Potential bias in the dataset due to self-selection of participants and reliance on video annotations.

Limitations

The dataset is modest-sized and self-collected, which may limit the generalizability of the findings.

Participant Demographics

294 subjects with a mean age of 68 years, including 97 females, with 150 healthy individuals and 144 diagnosed with Parkinson's Disease.

Statistical Information

P-Value

0.87

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1371/journal.pone.0315453

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication