Improvement of Pyramidal Tract Side Effect Prediction Using a Data-Driven Method in Subthalamic Stimulation

Area of Science:

• Neurosurgery
• Neurology
• Biomedical Engineering

Background:

• Subthalamic nucleus deep brain stimulation (STN DBS) is a key treatment for Parkinson's disease.
• A significant limitation of STN DBS is the occurrence of pyramidal tract side effects (PTSE).
• Accurate prediction of PTSE is crucial for surgical planning and patient safety.

Purpose of the Study:

• To compare the efficacy of two PTSE prediction methods in STN DBS.
• To evaluate a volume of tissue activated (VTA) modeling approach.
• To assess a novel data-driven artificial neural network model (PyMAN).

Main Methods:

• Postoperative PTSE was assessed by two clinicians in 20 Parkinson's disease patients.
• Two prediction methods were evaluated: VTA modeling and PyMAN.
• PyMAN utilizes nonlinear correlation between PTSE current threshold and 3-D electrode coordinates.

Main Results:

• 1696 electroclinical tests were used for method development and comparison.
• The PyMAN method demonstrated significantly higher sensitivity, specificity, and predictive values.
• Statistical analysis (Mann-Whitney U test) confirmed PyMAN's superior performance (P < 0.05).

Conclusions:

• The PyMAN method is more effective than VTA-based modeling for predicting PTSE.
• This data-driven tool can assist neurosurgeons in preplanning electrode trajectories.
• Improved PTSE prediction enhances surgical safety and optimizes STN DBS outcomes.