Back

One side effect - two networks? Lateral and postero-medial stimulation spreads induce dysarthria in subthalamic deep brain stimulation for Parkinsons Disease

Jergas, H.; Petry-Schmelzer, J. N.; Hannemann, J. H.; Thies, T.; Strelow, J. N.; Rubi-Fessen, I.; Quinting, J.; Baldermann, J. C.; Muecke, D.; Fink, G. R.; Visser-Vandewalle, V.; Dembek, T. A.; Barbe, M. T.

2023-04-28 neurology
10.1101/2023.04.26.23289100 medRxiv
Show abstract

ObjectiveThis study aims to shed light on structural networks associated with stimulation-induced dysarthria (SID) and to derive a data-driven model to predict SID in patients with Parkinsons Disease (PD) and deep brain stimulation (DBS) of the subthalamic nucleus (STN). MethodsRandomized, double-blinded monopolar reviews determining SID thresholds were conducted in 25 patients with PD and STN-DBS. A fiber-based mapping approach, based on the calculation of fiber-wise Odds Ratios for SID, was employed to identify the distributional pattern of SID in the STNs vicinity. The ability of the data-driven model to classify stimulation volumes as "causing SID" or "not causing SID" was validated by calculating receiver operating characteristics (ROC) in an independent out-of-sample cohort comprising 14 patients with PD and STN-DBS. ResultsLocal fiber-based stimulation maps showed an involvement of fibers running lateral and postero-medial to the STN in the pathogenesis of SID, independent of the investigated hemisphere. ROC-analysis in the independent out-of-sample cohort resulted in a good fit of the data-driven model for both hemispheres (AUCleft = 0.88, AUCright = 0.88). InterpretationThis study reveals an involvement of both, cerebello-thalamic fibers, as well as the pyramidal tract, in the pathogenesis of SID in STN-DBS. The results may impact future postoperative programming strategies to avoid SID in patients with PD and STN-DBS.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.