Estimation of intra-operative brain shift based on constrained Kalman filter

Area of Science:

• Neurosurgery
• Medical Imaging
• Computational Anatomy

Background:

• Neuronavigation systems are crucial for surgical precision.
• Brain shift, a deformation of the brain during surgery, reduces neuronavigation accuracy.
• Accurate estimation of brain shift is essential for improving surgical outcomes.

Purpose of the Study:

• To develop an improved method for estimating brain shift to enhance neuronavigation accuracy.
• To introduce a computationally efficient algorithm for real-time brain shift estimation.

Main Methods:

• Brain shift is modeled as a Gaussian random vector with known mean and unknown covariance.
• Brain surface imaging and linear elastic models are utilized.
• A constrained Kalman filter (CKF) is employed for estimation.
• A recursive constrained Kalman filter (RCKF) is proposed, avoiding large matrix inversions.

Main Results:

• Simulation results validate the theoretical framework.
• The proposed RCKF method demonstrates superior performance compared to an existing method.
• RCKF significantly reduces computational cost, making it suitable for operating room use.

Conclusions:

• The developed RCKF method effectively estimates brain shift.
• This approach enhances the accuracy and reliability of neuronavigation systems.
• The computational efficiency of RCKF facilitates its clinical application.