An Inertial Sensor-Based System for Synchronous Upper Extremity Kinematic Reconstruction and Neural Recordings During Awake Deep Brain Stimulation. (16th November 2020)

• Record Type: Journal Article
• Authors: Cajigas, Iahn; Diaz, Anthony; Prins, Noeline; Tan, Sze K; Prasad, Abhishek; Luca, Corneliu; Jagid, Jonathan R
• Is Part Of: Neurosurgery. Volume 67(2010)Supplement 1
• Publication Date: 2020-11-16
• Subjects: Nervous system -- Surgery -- Periodicals; 617.48005
• DOI: 10.1093/neuros/nyaa447_652
• Languages: English
• ISSNs: 0148-396X

Abstract

Abstract: INTRODUCTION: Recent advancements in the semiconductor industry have resulted in significant improvements to low cost wearable sensors. Inertial measurement unit (IMU) sensors are unique in their ability to capture movement information non-invasively. When coupled with adequate calibration and advanced signal processing algorithms, position and orientation of objects attached to IMUs can be reconstructed in real-time. Multiple IMU sensors can thus be used intraoperatively as an objective tool for assessing tremor and to study the relationship between single neurons or populations of neurons with coordinated multi-joint upper extremity movements. METHODS: A single male patient with essential tremor scheduled for VIM-DBS surgery on the right hemisphere was recruited for this pilot study. Two IMU sensors (IMeasureU Sensor, Vicon) were placed; one at the midpoint of the left forearm and the other at the midpoint of the upper arm. Intraoperatively, the IMU sensors were synchronized with the microelectrode recording (MER) system. Detailed passive and active arm movements were recorded during MER at various depths along the planned trajectory and during monopolar review with the DBS electrode. Reconstruction of the position and orientation of the two arm segments was performed off-line. Intracranial MER signals were analyzed for neural spiking activity using the open-source neural spike train analysis toolbox (nSTAT) for Matlab and then synchronized with the reconstructions. RESULTS: Analysis of movement-synchronized MER signals allowed identification of individual neurons at different depths along the planned trajectory. For example, at 3.2 mm above the VIM target, two morphologically distinct neural spike patterns were observed; one active during wider passive arm movements and the other during narrower movements. Using reconstructed movements, changes in arm tremor frequency for various stimulation parameter combinations were quantified, allowing for clear distinction between optimal vs suboptimal parameter combinations. CONCLUSION: With the benefit of being increasingly accurate, small, and inexpensive, IMU sensors can optimize the intraoperative evaluation of tremor and provide a way to correlate distinct neural firing patterns with individual extremity movements resulting in improved understanding of the receptive fields of neurons encountered during DBS trajectories.