Classification of older adults with/without a fall history using machine learning methods.

Area of Science:

• Gerontology
• Biomedical Engineering
• Computer Science

Background:

• Falls pose a significant health risk for the elderly population.
• Accurate fall risk assessment is crucial for effective prevention strategies.

Purpose of the Study:

• To develop and evaluate machine learning models for classifying older adults into high and low fall risk groups.
• To identify key gait features predictive of fall risk.

Main Methods:

• Utilized a 3D motion capture system to extract gait features from older adults.
• Applied and compared several machine learning algorithms: K-Nearest Neighbour, Naive Bayes, Logistic Regression, Neural Network, and Support Vector Machine.
• Assessed classification accuracy, sensitivity, and specificity for each model.

Main Results:

• Machine learning classifiers demonstrated capability in distinguishing between high and low fall risk groups.
• Feature extraction and algorithm tuning were critical for optimizing predictive performance.
• Various algorithms showed promising results in predicting fall risk based on gait analysis.

Conclusions:

• Machine learning offers a viable approach for objective fall risk assessment in older adults.
• Gait analysis combined with machine learning can enhance the prediction and prevention of falls.
• This study provides a foundation for developing advanced, personalized fall prevention interventions.