Fall Risk Prediction in Multiple Sclerosis Using Postural Sway Measures: A Machine Learning Approach

Abstract

Numerous postural sway metrics have been shown to be sensitive to balance impairment and fall risk in individuals with MS. Yet, there are no guidelines concerning the most appropriate postural sway metrics to monitor impairment. This investigation implemented a machine learning approach to assess the accuracy and feature importance of various postural sway metrics to differentiate individuals with MS from healthy controls as a function of physiological fall risk. 153 participants (50 controls and 103 individuals with MS) underwent a static posturography assessment and a physiological fall risk assessment. Participants were further classified into four subgroups based on fall risk: controls, low-risk MS (n = 34), moderate-risk MS (n = 27), high-risk MS (n = 42). Twenty common sway metrics were derived following standard procedures and subsequently used to train a machine learning algorithm (random forest – RF, with 10-fold cross validation) to predict individuals’ fall risk grouping. The sway-metric based RF classifier had high accuracy in discriminating controls from MS individuals (>86%). Sway sample entropy was identified as the strongest feature for classification of low-risk MS individuals from healthy controls. Whereas for all other comparisons, mediolateral sway amplitude was identified as the strongest predictor for fall risk groupings.

Nonlinear Measures to Evaluate Upright Postural Stability: A Systematic Review

“…in all of the cases and resulted in losing one point. Only three articles [ 50 , 52 , 55 ] were rated lower (9/11 points). In this case, a negative answer to question 10 (Have actual probability values been reported (e.g., 0.035 rather than <0.05) for the main outcomes except where the probability value is less than 0.001?)…”

“…Therefore, in this case of older adults lower SampEn represent more regular pattern for CoP variability, that is the rigid strategy, and lower adjustment to perturbations [ 85 , 86 ]. The comparison of healthy people with those suffering from chronic cervical pain, multiple sclerosis, cerebral palsy, chronic cervical pain following whiplash injury, or fibromyalgia [ 43 , 45 , 48 , 52 , 55 ] revealed that people with dysfunctions, injures, or diseases had lower values of entropy. Therefore, in this case the results of the entropy analysis suggest that people from this group need to concentrate more on postural control or else they lose complexity and automated postural control.…”

Balance Testing in Multiple Sclerosis—Improving Neurological Assessment With Static Posturography?

“…There have been numerous successful attempts to assess postural instability in MS patients (12,23,24), some of them to determine risk of falling (25)(26)(27) or to establish correlation with disease disability (26)(27)(28)(29). A better sensitivity for static posturography than the classical Romberg test has been previously reported, even with a possible prognostic value (24).…”

“…Static posturography involves the electronic evaluation of the body's center of pressure or gravity, recording a wide range of more than 100 balance-relevant parameters including speed, sway, root mean square distance, delineated area or 95% confidence ellipse as well as other values (7). Even though the choice of the ideal static posturography outcome measure could be problematic due to the immense amount of available variables (30), static outcomes equivalent to delineated area, average sway and average speed of sway calculated from mediolateral sway amplitude have been shown to be the strongest predictors to discriminate impaired people with MS from healthy subjects according to the results of a machine learning approach (25).…”

Development and Validation of Wearable Inertial Sensor System for Postural Sway Analysis

“…It was found that the application of stochastic resonance to children with CP significantly decreased their postural sway (Zarkou, Lee, Prosser, Hwang, & Jeka, 2018). In a randomized controlled trial to health care professionals with the experimental group given 8weeks of stochastic resonance whole-body vibration training and the control group with no intervention (Elfering et al, 2014), it was found that the mediolateral sway was significantly decreased in the experimental group but not in the control group, indicating that it may be beneficial in preventing balance-related injuries at work (Elfering et al, 2014), as medial-lateral sway amplitude is considered strong predictor of fall risk (Sun, Hsieh, & Sosnoff, 2018). The aim of this study was to develop and test the determinism of sway signals of a new wearable sensor device developed in our laboratory (Mini-Logger) to the gold standard camera-based motion capture system with specialized sway platform with 4 Degrees of Freedom (DOF).…”