An exergame system based on force platforms and body key-point detection for balance training

• Main Authors: Lavarda, Marcos D., de Borba, Pedro A., Oliveira, Matheus R., Borba, Gustavo B., de Souza, Mauren A., Gamba, Humberto R.
• Format: Conference Proceeding
• Language: English
• Subjects: Computers; Force; Games; Image segmentation; Load modeling; Protocols; Training
• ISSN: 1557-170X; 2694-0604
• DOI: 10.1109/EMBC.2016.7590636

Postural instability affects a large number of people and can compromise even simple activities of the daily routine. Therapies for balance training can strongly benefit from auxiliary devices specially designed for this purpose. In this paper, we present a system for balance training that uses the metaphor of a game, what contributes to the motivation and engagement of the patients during a treatment. Such approach is usually named exergame, in which input devices for posturographic assessment and a visual output perform the interaction with the subject. The proposed system uses two force platforms, one positioned under the feet and the other under the hip of the subject. The force platforms employ regular load cells and a microcontroller-based signal acquisition module to capture and transmit the samples to a computer. Moreover, a computer vision module performs body key-point detection, based on real time segmentation of markers attached to the subject. For the validation of the system, we conducted experiments with 20 neurologically intact volunteers during two tests: comparison of the stabilometric parameters obtained from the system with those obtained from a commercial baropodometer and the practice of several exergames. Results show that the proposed system is completely functional and can be used as a versatile tool for balance training.