Localization of the subthalamic nucleus in Parkinson disease using multiunit activity

Area of Science:

• Neurosurgery
• Neurology
• Biomedical Engineering

Background:

• Subthalamic nucleus (STN) localization is crucial for Parkinson disease (PD) deep brain stimulation (DBS) surgery.
• Current methods rely on qualitative intraoperative microelectrode recordings (MER), lacking objective precision.

Purpose of the Study:

• To prospectively validate a quantitative detection algorithm (QD) for STN localization using multi-unit activity (MUA).
• To compare the accuracy and precision of QD against traditional qualitative MER (IOM).

Main Methods:

• Ten PD patients undergoing STN DBS surgery were studied.
• Multi-unit activity (MUA) was processed using wavelet denoising and spectral analysis (500-2000Hz band).
• Quantitative detection (QD) results were compared to intraoperative microelectrode recording mapping (IOM) using Bland-Altman and Pearson's correlation analyses.

Main Results:

• The QD demonstrated strong correlation with IOM for both dorsal (r=0.79, p<0.0001) and ventral (r=0.91, p<0.0001) STN borders.
• Mean differences between IOM and QD were minimal (0.31±0.84mm dorsal, 0.44±0.47mm ventral).
• Only 5.6% of measurements fell outside the 95% confidence interval, indicating high agreement.

Conclusions:

• The quantitative detection algorithm (QD) provides objective and precise STN border localization.
• QD shows excellent agreement and strong correlation with the established intraoperative microelectrode recording (IOM) method.
• Further blinded comparative studies are warranted to confirm QD's superiority in guiding DBS lead placement.