Abstract

LUENGAS C, Lely A; CAMARGO CASALLAS, Esperanza  and  GARZON, Enrique Yamid. Evaluation of static postural stability measures using cluster. Ing. compet. [online]. 2023, vol.25, n.3, e-21512866.  Epub Sep 30, 2023. ISSN 0123-3033.  https://doi.org/10.25100/iyc.v25i3.12866.

Loss of somatosensory feedback in below-the-knee (transtibial) amputees implies a series of changes in the static standing posture, which leads to the affectation of the behavior of the center of pressure (CoP). The performance of two conventional CoP measures used for the characterization of static postural stability (SPS) is validated using unsupervised machine learning algorithms, applied to two population groups: the control group corresponds to non-amputee subjects and the amputee group to subjects with transtibial amputation. Scenarios are required for each of the algorithms using information theory as a classification method, data normalization is performed through binning. In the two measurements (velocity and displacement) two groups were identified, corresponding to the groups examined. A significant difference was observed between the groups, particularly in the CoP velocity, is the best discriminating variable. This study allows professionals interested in the subject to be guided about the variable to be used when analyzing the SPS, as well as making use of the datasets to support the prosthesis alignment part.

Keywords : Clustering; amputee; machine learning; information theory.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )