EQuilibrium: A Prototype of a Sensor-Based Balance Training and Monitoring System

Abstract

Abstract Integrated sensor devices are becoming a valuable tool for monitoring, evaluation, and real-time feedback during various exercises in sports, physical rehabilitation, and injury prevention. In this paper, we present the design and development of an inertial sensor system integrated into a fitness ball called eQuilibrium. The main application of this system would be balance training, analysis, and monitoring. The sensor system is composed of an inertial measurement unit sensor with a microcontroller and a WiFi communication module. The wireless sensor device is sending data to LabView application running on a laptop computer. The gyroscope and accelerometer provide a variety of signals, and we defined and evaluated three possible metrics to match the signals with the actual execution of the exercises. The metrics include i) angle rotation range, ii) mean absolute angular velocity, mean absolute acceleration deviation, and iv) mean absolute acceleration gradient. For the initial evaluation of the system, we specified test protocol and tasks that were executed by test subjects and monitored by a trained tester. There are four tasks in the protocol, and each task was being executed for at least 5 seconds. The first task was a reference task, standing on the feet of both legs. In the remaining tasks, the subject was standing on the tips of the toes (both legs, single leg), with the heels as high as possible off the ground. We conducted initial tests to analyze the operation of the system and obtain the initial measurement results. Our preliminary results indicate that several scalar metrics derived from motion sensor data can be used to evaluate motion intensity in the control of body balance. Rotation angle range, mean absolute rotation speed, mean absolute acceleration standard deviation, and mean absolute acceleration gradient are scalar values and independent of sensor axis orientation. In the future, we plan to perform more test executions and compare the sensor results with subjective self-reports of the test subjects, as well as the physical rehabilitation expert evaluations.