A Multi-Modular System for the Visualization and Classification of MER Data During Neurostimulation Procedures

Area of Science:

• Neurosurgery
• Medical Imaging
• Computational Neuroscience

Background:

• Accurate electrode placement is critical for effective neurostimulation therapy.
• Real-time feedback during surgery can significantly improve procedural accuracy.
• Integrating diverse data streams aids surgical decision-making.

Purpose of the Study:

• To develop an interactive analysis and visualization tool for improving electrode placement accuracy.
• To assist surgeons in real-time tissue type detection during neurostimulation surgery.
• To enhance surgical precision through fused visualization and AI-driven analysis.

Main Methods:

• Real-time processing of microelectrode recordings (MER).
• Utilizing a convolutional neural network (CNN) for tissue type classification.
• Integrating 2D/3D visualizations (CT, MRI, DVR, slice views) with MER data and CNN results.

Main Results:

• The system provides fused visualizations of electrode location, MER, and tissue classification.
• Real-time MER analysis and CNN output are presented in waveform and frequency domains.
• Pre-operative imaging (CT/MRI) is integrated for comprehensive spatial context.

Conclusions:

• The developed tool enhances surgical accuracy by providing real-time, multi-modal feedback.
• Interactive visualization and AI-driven tissue classification support critical intraoperative decisions.
• This system offers a significant advancement in precision for neurostimulation therapy.