Clinical frailty syndrome assessment using inertial sensors embedded in smartphones

Abstract

The aim of this study was to identify the series of kinematic variables demonstrating the greatest precision in discriminating between the function of two groups of elderly persons (frail and non-frail) in the 10 m expanded timed up and go (ETUG) test using inertial sensors embedded in the iPhone 4®. A cross-sectional study was conducted to identify the kinematic variables with the highest degree of precision in discriminating between the two groups. The predicted capability of the kinematic variables was evaluated using receiver operating characteristic curves. The sample comprised 30 participants over 65 years old, 14 frail and 16 non-frail, assessed for frailty syndrome using the Fried criteria. Acceleration variables discriminated between the participant groups in the study; specifically these were the peak negative acceleration variables for motion axes x, y and z. In terms of sensitivity, the values were greater than or equal to those for the variable traditionally used to discriminate in the ETUG test, namely time. The kinematic parameters obtained from the internal inertial sensors in the iPhone 4® are promising additions to the ETUG analysis. There are encouraging signs that the analyses of these parameters in the separate phases of the ETUG procedure offer the potential for improved discrimination between frail and non-frail individuals. However, further in-depth study is required to verify the findings.