We have developed a wireless wearable stimulation system that analyzes upper limb tremor using a three-axis accelerometer and that modulates/attenuates tremor using peripheral-nerve electrical stimulation with adjustable stimulation parameters and a real-time tremor detection algorithm.
“Actual experimental setup for the bean transfer tasks: One of the subjects performed the bean-transfer task while wearing the wrist device and electrodes..”
Overview of Longitudinal Tremor Monitor and Stimulation System (LTMS) System: (1) effects of stimulation parameters on tremor movements and (2) clinical efficacy of the peripheral-nerve electrical stimulation with parameter optimization for ADLs.
Wearable Neuromodulation Hardware
Details of the wearable tremor-modulation system. The system consists of a wearable wrist device that contains an accelerometer, an electric stimulation unit, a 2.4 GHz transceiver, and an off-board signal processing system linked to the wearable device. (a) Off-board signal processing flow for tremor detection and parameter optimization, (b) the voltage mode stimulator (biphasic stimuli are generated by alternating between turning on the electronic switches 1 and 1, (c) the wearable wrist device, (d) experimental apparatus, including the wrist device and a pair of stimulating surface electrodes, (e) custom-designed printed circuit boards for the wrist device, (f) overall data flow from a PC or smartphone to the wrist device, (g) graphical user interface for the experimental setup (i.e., stimulation threshold setting and a phase-locked stimulation trial), and (h) a custom-designed wireless transceiver for the wrist device.
Tremor recording obtained during a control trial (stimulation OFF, a, c, and e) and a stimulation trial (stimulation ON, b, d, and f) for participant ET04. Raw three-axis accelerometry data obtained while the patient performed the bean-transfer task are shown in (a) and (b), while (c) and (d) show results of the HPF-AQM output of the same data. The plots in (e) and (f) show spectrograms (window: 1/2 sampling frequency (Fs), overlap: Fs) and the average power density for the data shown in (c) and (d).
Effects of peripheral nerve stimulation over the course of trials. Overall tremor movement changes without and with stimulation for (a) dominant tremor frequency and (b) tremor power. The error bar represents the standard error; ⋆: P < 0.05; ∗: P < 0.005.
We are currently working on the parameter optimization of the upper-limb neuromodulation to subside the tremor movement in real time.