Localization of Target Structures through Data Fusion Applied to Neurostimulation

Area of Science:

• Neurosurgery
• Medical Imaging
• Computational Neuroscience

Background:

• Parkinson's Disease symptoms are often managed with Deep Brain Stimulation (DBS).
• Accurate electrode placement within specific brain structures like the Subthalamic Nucleus (STN) and Internal Globus Pallidus (GPi) is critical for DBS efficacy.
• Current methods for identifying these deep brain targets can be time-consuming and require significant expertise.

Purpose of the Study:

• To develop an automated method for identifying deep brain structures essential for DBS surgery.
• To enhance the accuracy and efficiency of surgical planning for Parkinson's Disease treatment.

Main Methods:

• Utilized data fusion combining anatomical Magnetic Resonance Imaging (MRI) with expert anatomical knowledge.
• Employed a fuzzy logic approach within a possibilist framework to manage information.
• Developed a multi-scale graph-based virtual atlas to model anatomical knowledge.
• Matched the virtual atlas to patient MRI data to estimate target structure locations.

Main Results:

• The automated method demonstrated promising results in identifying deep brain structures on test images.
• The approach successfully estimated the location of key targets like the STN and GPi.
• Physician assistance in target definition was supported by the method's output.

Conclusions:

• An automated, data-fusion-based method shows potential for accurate identification of DBS targets.
• This approach can assist neurosurgeons in defining optimal electrode placement for Parkinson's Disease.
• The long-term goal is to enable fully automated surgical planning for DBS procedures.