Data-guide for brain deformation in surgery: comparison of linear and nonlinear models

Area of Science:

• Neurosurgery
• Medical Imaging
• Computational Mechanics

Background:

• Pre-operative imaging provides high-resolution brain data, but intra-operative imaging is low-resolution.
• Accurate surgical navigation requires deforming pre-operative images to match intra-operative brain geometry.
• Brain shift after craniotomy introduces significant localization errors.

Purpose of the Study:

• To develop and evaluate a nonlinear biomechanical model for predicting brain deformation during surgery.
• To compare the accuracy of the nonlinear model against existing linear models.
• To assess the model's ability to localize brain regions using intra-operative data.

Main Methods:

• Employing biomechanical models guided by low-resolution intra-operative images.
• Utilizing finite element methods to solve differential equations governing brain deformation.
• Optimizing model parameters by identifying corresponding points between pre- and intra-operative images.

Main Results:

• The nonlinear model demonstrated reduced localization error caused by brain deformation.
• Evaluation using simulated and real data showed superior performance of the nonlinear model.
• Partial validation with intra-operative images confirmed the model's predictive capabilities.

Conclusions:

• The proposed nonlinear model offers more accurate predictions of brain deformation than linear models.
• The model can predict the deformation of the entire brain using limited intra-operative surface data.
• The computational execution time of the nonlinear model is significantly longer than linear models.