High density microelectrode recording predicts span of therapeutic tissue activation volumes in subthalamic deep brain stimulation for Parkinson disease

Abstract

Background: Subthalamic deep brain stimulation alleviates motor symptoms of Parkinson disease by activating precise volumes of neural tissue. While electrophysiological and anatomical correlates of clinically effective electrode sites have been described, therapeutic stimulation likely acts through multiple distinct neural populations, necessitating characterization of the full span of tissue activation. Microelectrode recordings have yet to be mapped to therapeutic tissue activation volumes and surveyed for predictive markers. Objective: Combine high-density, broadband microelectrode recordings with detailed computational models of tissue activation to describe and to predict regions of therapeutic tissue activation. Methods: Electrophysiological features were extracted from microelectrode recordings along 23 subthalamic deep brain stimulation implants in 16 Parkinson disease patients. These features were mapped in space against tissue activation volumes of therapeutic stimulation, modeled using clinicallydetermined stimulation programming parameters and fully individualized, atlas-independent anisotropic tissue properties derived from 3T diffusion tensor magnetic resonance images. Logistic LASSO was applied to a training set of 17 implants out of the 23 implants to identify predictors of therapeutic stimulation sites in the microelectrode recording. A support vector machine using these predictors was used to predict therapeutic activation. Performance was validated with a test set of six implants. Results: Analysis revealed wide variations in the distribution of therapeutic tissue activation across the microelectrode recording-defined subthalamic nucleus. Logistic LASSO applied to the training set identified six oscillatory predictors of therapeutic tissue activation: theta, alpha, beta, high gamma, high frequency oscillations (HFO, 200e400 Hz), and high frequency band (HFB, 500e2000 Hz), in addition to interaction terms: theta x HFB, alpha x beta, beta x HFB, and high gamma x HFO. A support vector classifier using these features predicted therapeutic sites of activation with 64% sensitivity and 82% specificity in the test set, outperforming a beta-only classifier. A probabilistic predictor achieved 0.87 area under the receiver-operator curve with test data. Conclusions: Together, these results demonstrate the importance of personalized targeting and validate a set of microelectrode recording signatures to predict therapeutic activation volumes. These features may be used to improve the efficiency of deep brain stimulation programming and highlight specific neural oscillations of physiological importance.

Randomized, Double-Blind Assessment of LFP Versus SUA Guidance in STN-DBS Lead Implantation: A Pilot Study

“…Interestingly, we also noted HFOs above STN in three patients (see Figure 2E). This activity could be originating from other structures such as thalamus or zona incerta (ZI) (Thompson et al, 2014;Yang et al, 2014;Lu et al, 2019;Meidahl et al, 2019). Previous work has shown that dorsolateral STN and ZI stimulation provides the greatest improvement in PD motor symptoms (Gourisankar et al, 2018), which correlates with our observation.…”

“…However, the volatile interface stability and increased computational power requirement arising from higher sampling rates might still favor LFPs (Rouse et al, 2011;Buzsáki et al, 2012;Priori et al, 2013;Thompson et al, 2014). Growing literature supports the utility of LFPs in intraoperative mapping (Chen et al, 2006;Przybyszewski et al, 2016;Telkes et al, 2016;Kolb et al, 2017;Lu et al, 2019). Our results also support the use of LFPs intraoperatively for DBS lead implantation.…”

“…There is an abundance of studies reporting the response of LFP-derived biomarkers to medication (Foffani et al, 2003;Priori et al, 2004;Marceglia et al, 2006;Kane et al, 2009;Lopez-Azcarate et al, 2010;Özkurt et al, 2011;Ozturk et al, 2019) and DBS (Kühn et al, 2008;Eusebio et al, 2011;McConnell et al, 2012) therapies, as well the correlation between these biomarkers and cardinal symptoms of PD (Kühn et al, 2006;Weinberger et al, 2006;Ray et al, 2008;Lopez-Azcarate et al, 2010;Oswal et al, 2013;Brittain and Brown, 2014;Ozturk et al, 2019). We and others have previously shown that these patterns can provide utility in contact selection (Ince et al, 2010;Connolly et al, 2015) or targeting the optimal location for DBS implantation (Chen et al, 2006;Thompson et al, 2014;Telkes et al, 2016;Kolb et al, 2017;Lu et al, 2019). Specifically, oscillations in the beta and HFO range and their cross-frequency interactions have been used to pinpoint the "sweet spot" for DBS (Wang et al, 2014;Connolly et al, 2015;Telkes et al, 2016;Horn et al, 2017;van Wijk et al, 2017;Hell et al, 2019).…”

Machine Learning’s Application in Deep Brain Stimulation for Parkinson’s Disease: A Review

“…High accuracy region SVM classification models were used by Guillén et al [92,93] with a 99.5% accuracy at stratifying MERs between the STN, thalamus nucleus, and the substantia nigra. Similarly, Lu et al [94] applied SVM to 16 subjects' MERs guided by 3-Tesla MRIs. This model achieved an AUC of 0.87.…”

Mapping of subthalamic nucleus using microelectrode recordings during deep brain stimulation

“…Several previous studies have shown the association of dorsolateral posterior STN field potentials and firing rate to effective clinical outcome 13 – 15 , quantification of motor subtypes 16 , severity of rigidity and bradykinesia in PD 17 , optimal DBS implantation trajectory 18 and simulation parameters 19 among others. Recent studies have further used various recordings using MER for localizing dorsal–ventral border of STN 20 and predicting therapeutic volume of tissue activation 21 . Despite extensive research in the pathophysiological role and clinical correlation of these electrophysiological activity patterns in PD and target localization, studies focusing on the differential use of them for improving the accurate implantation of the electrodes in sensorimotor region of STN is still insufficient.…”