System for Postural Stability Assessment Based on an Inertial Sensors Network

This paper presents a system for quantitative analyzing the postural stability and balance of elderly people. For doing that, a computational biomechanics model is developed, for which body segment information from lower limps are captured by using an inertial sensors network. For this, an alignment and calibration strategy based on quaternion theory is implemented, which constructs a reference coordinate system aligned and calibrated to each body segments for obtaining the joint movements from elderly in static state. Once biomechanics model is calibrated, the body centre of mass is estimated by approximating the origin of the sensor from the distal to proximal body segments, which allows the segmentation of each anatomical region of the body. Finally, the complete body centre of mass from biomechanics model is obtained for postural balance analysis. The proposed system was evaluated with volunteers using 3 different stability tests, the data obtained in each stability test were analyzed with box plots and a K-means classifier, establishing indices of postural stability and showing alterations during the testing phase. The results demonstrate some difficult levels to maintain postural balance during these stability test, where postural balance on one foot proved to be a great challenge for elderly people, with displacements of their centre of mass of around 30 cm. The contribution of this work is to analyze the deterioration of stability in elderly people based on the centre of mass (COM), which in the future could be used as a study alternative for specialized walking centers.