Functional localization and visualization of the subthalamic nucleus from microelectrode recordings acquired during DBS surgery with unsupervised machine learning

Abstract

Microelectrode recordings are a useful adjunctive method for subthalamic nucleus localization during deep brain stimulation surgery for Parkinson's disease. Attempts to quantitate and standardize this process, using single computational measures of neural activity, have been limited by variability in patient neurophysiology and recording conditions. Investigators have suggested that a multi-feature approach may be necessary for automated approaches to perform within acceptable clinical standards. We present a novel data visualization algorithm and several unique features that address these shortcomings. The algorithm extracts multiple computational features from the microelectrode neurophysiology and integrates them with tools from unsupervised machine learning. The resulting colour-coded map of neural activity reveals activity transitions that correspond to the anatomic boundaries of subcortical structures. Using these maps, a non-neurophysiologist is able to achieve sensitivities of 90% and 95% for STN entry and exit, respectively, to within 0.5 mm accuracy of the current gold standard. The accuracy of this technique is attributed to the multi-feature approach. This activity map can simplify and standardize the process of localizing the subthalamic nucleus (STN) for neurostimulation. Because this method does not require a stationary electrode for careful recording of unit activity for spike sorting, the length of the operation may be shortened.

Depth-Time Interpolation of Feature Trends Extracted from Mobile Microelectrode Data with Kernel Functions

“…We used a 4-second duration for T, and the following lengths for D: 0.025, 0.05, 0.075, 0.1, 0.15, and 0.2 mm. Our value for T was based on observations that indicated that this duration provides an empirically optimal trade-off between increased spatial resolution and feature stability at the electrode velocities we typically employed ( [14] , data not shown). For each window, we used the time and depth at the temporal center of the window as the time and depth stamp.…”

“…The computational features used here were reported previously [14,16] . We made a slight modification by normalizing each feature to the data window length (e.g.…”

“…Though this optimizes the traditional clinical procedure of microelectrode-based target localization, the resultant neurophysiological data are irregularly sampled in space because of (1) electrode slowing near critical areas and (2) electrode pauses during clinical testing. The automated feature output (typically extracted via temporal data windows in a sliding window paradigm) is similarly irregularly sampled, such that one must examine a temporal trend (time vs. feature activity), as well as a separate timeversus-depth plot [14] , to localize a target. This adds complexity to the localization task.…”

Microelectrode Recording Duration and Spatial Density Constraints for Automatic Targeting of the Subthalamic Nucleus

“…Typically, MER signals are observed both visually and with audio by the physiologist and/or neurosurgeon during surgery to identify the functional targets. Recently, several studies have shown that MER data can be utilized for automatic localization of the STN with reduced variation and better accuracy [25,26,27,28,29,30]. Commonly used signal features include the background noise level, spike count and power spectral density (PSD) in the beta and theta (tremor) frequency bands.…”

Symbolic dynamics for localization of the subcortical structures during deep brain stimulation surgery for Parkinson´s disease

“…The possibility of targeting errors to DBS necessitates the use of some form of intraoperative neurophysiologic monitoring to confirm the correct targeting during surgery. The purpose of the development of numerical techniques to MER processing is to assist the surgical team in determining the optimal location of the lesion or DBS lead [2][3][4].…”