Assessing Locomotive Syndrome Through Instrumented Five-Time Sit-to-Stand Test and Machine Learning.

Area of Science:

• Geriatric Medicine
• Biomedical Engineering
• Machine Learning in Healthcare

Background:

• Locomotive syndrome (LS) impairs daily activities, necessitating early detection to prevent nursing care needs.
• The Geriatric Locomotive Function Scale (GLFS-25) is a subjective tool for LS staging.
• Objective, technology-based assessments are needed to complement or replace subjective measures.

Purpose of the Study:

• To develop and validate a machine learning model for quantitative assessment of locomotive syndrome (LS) stages.
• To evaluate the efficacy of an instrumented five-time sit-to-stand test (FTSTS) for LS staging.
• To explore the potential of inertial measurement units (IMUs) and machine learning for objective LS assessment.

Main Methods:

• Participants performed an instrumented five-time sit-to-stand test (FTSTS) with a single pelvic inertial measurement unit (IMU).
• 144 features were extracted from acceleration data, and seven machine learning models were trained.
• The multilayer perceptron (MLP) model, enhanced with data augmentation and principal component analysis (PCA), was evaluated.

Main Results:

• The MLP+PCA model achieved high performance metrics: 0.9 accuracy, 0.92 precision, 0.9 recall, and 0.91 F1 score.
• This demonstrates the model's effectiveness in accurately classifying LS stages.
• The study highlights the potential of using IMUs and machine learning for objective LS assessment.

Conclusions:

• Machine learning analysis of FTSTS data using IMUs provides a highly accurate and objective method for assessing locomotive syndrome (LS).
• This approach offers a promising foundation for developing remote LS monitoring systems using accessible technology.
• Objective quantitative assessments can significantly aid in the early detection and management of locomotive syndrome.