Related Experiment Video
Updated: Jul 8, 2026
06:28
Screening People on Standing Balance with Romberg Testing and Walking Balance with Tandem Walking
Published on: September 1, 2023
Wavelet-based feature extraction for support vector machines for screening balance impairments in the elderly
Ahsan H Khandoker1, Daniel T H Lai, Rezaul K Begg
1Department of Electrical and Electronic Engineering, The University of Melbourne, Melbourne, Australia. a.khandoker@ee.unimelb.edu.au
Summary
This study shows wavelet analysis of minimum foot clearance (MFC) effectively identifies elderly fall risk. This advanced gait analysis using support vector machines (SVMs) offers superior accuracy for balance impairment screening.
Area of Science:
- Biomedical Engineering
- Gerontology
- Signal Processing
Background:
- Falls are a major health concern for the elderly, often linked to gait abnormalities.
- Early detection of gait impairments can significantly reduce fall-related injuries.
- Minimum Foot Clearance (MFC) is a key gait variable associated with fall risk.
Purpose of the Study:
- To compare wavelet-based multiscale analysis with histogram analysis of MFC for fall risk screening.
- To develop and evaluate a Support Vector Machine (SVM) model for classifying gait patterns in the elderly.
- To assess the effectiveness of wavelet features in identifying balance impairments.
Main Methods:
- Recorded MFC during treadmill walking in healthy elderly and those with a history of falls.
- Applied wavelet-based multiscale analysis to MFC time series.
- Extracted features from MFC histograms and wavelet decomposition.
- Utilized SVMs to classify gait patterns based on extracted features.
Main Results:
- Wavelet-based features achieved 100% classification accuracy in an SVM model.
- MFC histogram analysis yielded 86.95% accuracy with statistical features.
- SVM posterior probabilities were calculated for estimating relative fall risk.
Conclusions:
- Wavelet-based analysis of MFC demonstrates superior performance for detecting balance impairments in the elderly.
- This approach shows promise for screening individuals at risk of falls and evaluating interventions.
- The SVM model using wavelet features is a valuable tool for falls prevention strategies.
