Risk assessment methodology for trajectory planning in keyhole neurosurgery using genetic algorithms

Abstract

Background: Preoperative assessment to find the safest trajectory in keyhole neurosurgery can reduce post operative complications.

Methods: We introduced a novel preoperative risk assessment semiautomated methodology based on the sum of N maximum risk values using a generic genetic algorithm for the safest trajectory search.

Results: A set of candidates trajectories were found for two surgical procedures. The trajectories search is done using a risk map considering the proximity of voxels within risk structures in multiple points and a genetic algorithm to avoid an exhaustive search. The trajectories were validated by a group of neurosurgeons.

Conclusions: The trajectories obtained with the proposal method were shorter in 5% and have greater distance from the voxels within the blood vessels in 4.7%. The use of genetic algorithm (GA) speeds up the search for the safest trajectory, decreasing in 99.9% the time required for an exhaustive search.

Keywords: genetic algorithms; keyhole neurosurgery; risk assessment; robotic surgery; trajectory planning.