Objective: To evaluate clinical usability of LFP(local field potentials)-physiomarkers and lead localisation for selection of suitable DBS electrode configurations.

Background: Selection of optimal contacts in DBS programming has become increasingly challenging due to technical innovations with multiple (segmented) contacts and independent current sources. Narrowing the search-space due to a ranking of availible contacts can speed up clinical programming considerably. Pathologic beta-band synchronisation is a well established physiomarker that correlates with akinetic-rigid symptoms.

Method: LFP were recorded from externalised DBS leads in patients undergoing DBS with 8-contact-leads (Boston Sc, linear/directional) for advanced PD. Recordings for resting-state and self-paced handmovements and MRI/CT imaging data were analysed in Matlab with SPM12, Fieldtrip and Lead-DBS-toolboxes, respectively. At rest beta-power within total beta-band (13-30Hz), low (13-20Hz) and high-beta-range (20-30Hz), HFO (200-400Hz) and phase reversal within beta-band were compared. During self-paced movements event-related desynchronisation (ERD) in the beta-range was quantified. All contacts were scored and ranked based on the respective physiomarker or distance to sweet spot as determined by Lead-DBS. Rankings were then compared to clinical real-world outcomes for parameter settings.

Results: Selecting the top 3 contacts for clinical stimulation based on ranking of contacts by LFP-power (AUC) in the entire beta-range was significantly better than selecting 3 arbitrary contacts (p<.05) for linear leads. The contact closest to the sweet spot as determined by Lead-DBS was significantly more likely to be among the top 3 beta-power contacts compared to a random contact (p<.05). HFO and beta-ERD were only detectable less than 50% of the recordings and did not show correlation to clinical settings.

Conclusion: Ranking based on power in entire beta-range showed the best correlation with DBS-parameter settings for linear leads. As shown in 4-contact-leads power in beta-range correlates with electrode-distance to the sweet spot. Other physiomarkers were either inconclusive or not robustly detectable. Also no robust marker for detection of directionality in directional leads could be established. Ranking of contacts by total beta-power and localisation is a promising method to set a starting point for clinical programming but still individual testing and clinical expertise remain the conerstone.

« Back to 2019 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/comparison-of-lfp-physiomarkers-for-dbs-programming/