24th EANN 2023, 14 - 17 June 2023, León, Spain

Towards Automatic Assessment of Quiet Standing Balance During the Execution of ADLs

Irene Guardiola-Luna, Leticia Monasterio-Expósito, Javier Macías-Guarasa, Álvaro Nieva-Suárez, Marina Murillo-Teruel, Jose Luis Martín-Sánchez, Sira Elena Palazuelos-Cagigas


  The current method used to estimate the balance a person has during the performance of Activities of Daily Life (ADLs) is through the application of standardized scales used by occupational therapists to evaluate a person's motor skills and performance quality during those activities, such as the Assessment of Motor and Process Skills scale (AMPS). In this paper, we propose a method to automate the evaluation of a person's balance during the stage of quiet standing still while a person is completing an ADL. Our proposal is aimed to first estimate the projection of the person's center of mass (CoM) from the 3D position of the body joints by applying theoretical and deep learning approaches. Then, we aim to predict a clinically validated objective balance score from previous estimations of the CoM and the Center of Pressure (CoP), using different neural network models. While there are other proposals in the literature, the lack of publicly available datasets makes it difficult to do an extensive comparison, so we compare our proposal with state-of-the-art results in two publicly available datasets, improving their results.  

*** Title, author list and abstract as seen in the Camera-Ready version of the paper that was provided to Conference Committee. Small changes that may have occurred during processing by Springer may not appear in this window.