Falls are a major problem not only for older adults, but also for health care systems, which have to spend enormous amounts of money in hospitalizations and long-term rehabilitation as result of injurious falls. We aim to assess the risk of falling as easy, accurate and inexpensive as possible to be able to identify those persons at high risk of functional decline and trigger appropriate clinical interventions. A new method are functional assessments with inertial sensors like the actibelt®. We analyzed the predictive value of the data recorded with this device in 171 community-dwelling female seniors with a mean age of 68 years included in the EU-Project VPHOP.
Results showed that acceleration-based parameters provide an improvement in risk prediction over classical features. We could prove 18 variables retrieved by the inertial sensor to be valid or plausible. To identify variables with predictive power, we applied the following feature selection and dimension reduction techniques: sparse Partial Least Squares (sPLS), Elastic Nets (EN) and Principal Components Analysis (PCA). We found that PCA is useless for this kind of data, whereas EN are most applicable according to our results. From a range of different classification methods, Logistic Regression (LR) and Support Vector Machines (SVM) showed the best performance. LR presented a higher specificity in the classification of fallers/non-fallers with a True Positive Rate (TPR), or rate of correctly classified fallers, equal to 0.30 and a True Negative Rate (TNR), or rate of correctly classified non-fallers, equal to 0.92. The results of SVM were more sensitive with TPR=0.60 and TNR=0.85. The latter results are more suitable because the consequences of incorrectly classifying a non-faller are less harmful than vice versa. The commonly used accuracy has been shown to be insufficient to describe the performance of classifiers on datasets with strongly differing class sizes. Instead, we suggest the use of Youden's J or the F-score.
«
Falls are a major problem not only for older adults, but also for health care systems, which have to spend enormous amounts of money in hospitalizations and long-term rehabilitation as result of injurious falls. We aim to assess the risk of falling as easy, accurate and inexpensive as possible to be able to identify those persons at high risk of functional decline and trigger appropriate clinical interventions. A new method are functional assessments with inertial sensors like the actibelt®. We...
»