Developing a Device- Independent Software for Adaptive Tremor Suppression Orthoses

Tremors are a debilitating symptom of common neurological diseases such as Parkinson's and Essential Tremor which affects over 10 million people in the U.S. alone. Currently, non-invasive tremor-suppression orthoses under research development are limited by their inability to calculate precise in-phase suppression signals and calibrated torques for controlled tremor suppression. Current solutions to this problem involve complex modeling of the orthosis device and require software remodeling when the hardware is changed. In addition, adaptation to personal conditions with various loads is not covered in the literature. To address these issues, a device-independent, adaptive tremor suppression library for tremor orthoses is developed. The library filters tremor-induced motions from a tremor patient's voluntary movements from sensor signals. It then predicts an inphase actuator response that suppresses tremor-induced motions in real-time. A deep neural network (CNN-LSTM) inferences a sequence of future tremor signals to account for processing and actuator response time delays. The library can calibrate the time delay precisely and convert the model output through a feedback loop into a calibrated torque signal, which an actuator uses to suppress a tremor with solely the necessary counteractive force. The library achieves tremor signal prediction by an average phase error of 2.8ms and an average amplitude suppression efficiency of 87% when predicting 50ms ahead on sequences of real patient data. This library serves as the first modular software for smart tremor suppression, and will later be implemented onto edge devices.