As people in developed countries live longer, assessing the fall risk becomes more important. A major contributor to the risk of elderly people falling is postural instability. This study aimed to use the multiscale entropy (MSE) analysis to evaluate postural stability during a timed-up-and-go (TUG) test. This test was deemed a promising method for evaluating fall risk among the elderly in a community. The MSE analysis of postural instability can identify the elderly prone to falling, whereupon early medical rehabilitation can prevent falls. Herein, an objective approach is developed for assessing the postural stability of 85 community-dwelling elderly people (aged 76.12 ± 6.99 years) using the short-form Berg balance scale. Signals were collected from the TUG test using a triaxial accelerometer. A segment-based TUG (sTUG) test was designed, which can be obtained according to domain knowledge, including “Sit-to-Walk (STW),” “Walk,” “Turning,” and “Walk-to-Sit (WTS)” segments. Employing the complexity index (CI) of sTUG can reveal information about the physiological dynamics’ signal for postural stability assessment. Logistic regression was used to assess the fall risk based on significant features of CI related to sTUG. MSE curves for subjects at risk of falling (n = 19) exhibited different trends from those not at risk of falling (n = 66). Additionally, the CI values were lower for subjects at risk of falling than those not at risk of falling. Results show that the area under the curve for predicting fall risk among the elderly subjects with complexity index features from the overall TUG test is 0.797, which improves to 0.853 with the sTUG test. For the elderly living in a community, early assessment of the CI for sTUG using MSE can help predict the fall risk.

The results obtained by medical experts and inertial sensors via clinical tests to determine fall risks are compared. A clinical test is used to perform the whole timed up and go (TUG) test and segment-based TUG (sTUG) tests, considering various cutoff points. In this paper, (a) t-tests are used to verify fall-risk categorization; and (b) a logistic regression with 100 stepwise iterations is used to divide features into training (80%) and testing sets (20%). The features of (a) and (b) are compared, measuring the similarity of each approach’s decisive features to those of the clinical-test results. In (a), the most significant features are the Y and Z axes, regardless of the segmentation, whereas sTUG outperforms TUG in (b). Comparing the results of (a) and (b) based on the overall TUG test, the Z axis multiscale entropy (MSE) features show significance regardless of the approach: expert opinion or logistic prediction. Among various clinical test combinations, the only commonalities between (a) and (b) are the Y-axis MSE features when walking. Thus, machine learning should be based on both expert domain knowledge and a preliminary analysis with objective screening. Finally, the clinical test results are compared with the inertial sensor results, prompting the proposal for multi-oriented data analysis to objectively verify the sensor results.