Novel methods of using accelerometry for upper limb FES control [conference paper]

Introduction: Accelerometry offers a low cost, low power solution to measuring body segment angle (BSA) relative to the gravity vector under quasi-static conditions. BSA may be used as an input to a state-machine based, upper limb functional electrical stimulation (FES) system. However, existing methods to obtain angle from the vertical all suffer from poor sensitivity when the sensitive axis approaches the vertical. This paper reports on two alternative methods (uncalibrated and calibrated) that use 3-axis accelerometer data to track body segment angle with respect to gravity. Methods: The uncalibrated method calculates the angle between the accelerometer x-axis and the gravity vector. The calibrated method uses a calibration rotation to define the measurement plane and the positive rotation direction. This method then calculates the component of rotation that is in the same plane as the calibration rotation. Both methods use an algorithm that switches between using sine and cosine, depending on the measured angle, which overcomes the poor sensitivity problem seen in previous methods. Results: A protractor system was used to test the accuracy of the algorithms. During the test, the protractor was moved from 0° to 180°, with short pauses at 30°, 60°, 90° and 150°. Maximum mean error for the uncalibrated method was 2.5° when the angle was near 0°. However, when testing the calibrated method, significant errors were observed when the measured angle was 150°. Discussion: The uncalibrated method is insensitive to rotation about the x-axis, a feature which could be advantageous for controlling upper limb FES if, for example, stimulation is to be triggered by lift of the forearm, but not as a result of pronation-supination. To do this, the x-axis of the accelerometer would be simply aligned with the long axis of the forearm. The calibrated method has the advantage that it can provide the sign of the angle change. Conclusion: Two novel methods for calculating BSA from 3-axis accelerometer data have been presented. The performance of the uncalibrated method led to it being incorporated into a flexible state-machine controller for the real-time control of FES during upper limb rehabilitation