The Study Of A Physical Model Of Brain Deformation During Neurosurgical Navigation

PartⅠConstruction & validation of linear elastic model of intraoperative brain deformation during craniotomy in swineOBJECTIVE:To compensate for the intraoperative experimental brain deformation for the purpose of increasing the accuracy of neurosurgical navigation.MATERIALS and METHODS:A linear elastic model of porcine brain based on finite element method(FEM) was constructed.After craniotomy in swine,the deformation of cortical surface was tracked by a 3D laser range scanner(LRS) as a boundary condition.This boundary condition was then applied on the finite element equation to simulate the entire brain deformation.Polymethyl methacrylate(PMMA) beads were implanted into the porcine brain as markers of tissue shift,and the predictive accuracy of the model was validated by the real-time data of brain deformation acquired by MR imaging during operation.RESULTS:The predictive error of this model ranged from 0.20 to 1.54mm,(mean 0.97±0.44mm).The accuracy of calibration ranged from 56.5%to 90.0%(mean 68.0±9.6%).The predictive accuracy on the displacement of superficial markers was higher than that of deeper markers(70.7±9.1%:65.4±10.8%,P＜0.05).CONCLUSIONS:Model-updated image was proved to be efficient,convenient,and reliable in animal research,therefore,it is an ideal approach to compensate for the brain deformation during neurosurgical navigation. PartⅡConstruction & validation of linear elastic model of intraoperative brain deformation during neurosurgical navigationOBJECTIVE:To compensate for the intraoperative human brain deformation for the purpose of increasing the accuracy of neurosurgical navigation.MATERIALS and METHODS:A linear elastic model of human brain based on finite element method(FEM) was constructed in 11 patients with craniotomies.After craniotomy,the deformation of cortical surface was tracked by a 3D laser range scanner(LRS) as a boundary condition which was applied on the finite element equation to simulate the entire brain deformation.The predictive accuracy of the model was validated by the real-time data of brain deformation acquired by an open intraoperative MR system.RESULTS:The predictive error of this model ranged from 1.29 to 1.91 mm(mean 1.62±0.22 mm).The accuracy of calibration ranged from 62.8%to 81.4%(mean 69.2±5.3%).The predictive accuracy on the displacement of the superficial structures was higher than that of the deeper structures(71.3%±6.1:66.8±5.0%,P＜0.01).The predictive accuracy of the tissue shift in central area of the bone window was similar to that in marginal area(72.9±7.3%:69.7±7.4%,P＞0.05).There was no significant difference on calibration accuracy between the sinking cortex group and the bulging cortex group(70.1±6.6%:68.1±3.7%,P＞0.05).Moreover,the calibration accuracy in the horizontal bone window group was higher than that in the non-horizontal bone window group(72.0±5.3%:65.7±2.9%,P＜0.05).CONCLUSIONS:Model-updated image was proved to be efficient,convenient,and reliable in clinical research,therefore,it is an ideal approach to compensate for the brain deformation during neurosurgical navigation.