The Development And Experimental Validation Of A Brain Deformation Compensation System For IGNS

The brain deformation after craniotomy mostly contributes to predictive error in image-guided neurosurgery (IGNS). In order to increase the accuracy of IGNS, we developed a brain deformation System(BDS) which is based on linear elastic model to correct brain deformation during neurosurgery. The device can incorporate the intra-operation data obtained by a laser range scanner(LRS) to track the brain surface on the region of craniotomy,compute the deformation field using the finite element method(FEM) and warp the pre-operative image.So a brain deformation framework named3Dimage system was designed and performed built on the principle above. Then the key technologies such as input and output, visualization, calibration, segmentation, mesh, surface tracking, FEM, warping the pre-operative image and network co mmunication were introduced and implemented.We conducted two validation experiments for the BDS. In the first experiment, we tested5pigs to validate the accuracy by estimating the displacements of mark points with this system. According to our experiments, the average shift error after craniotomy was reduced to0.71±0.14mm from1.80±0.30mm. It showed even encouraging results with respect to subsurface shift, the shift error1.79±0.67mm on average comparative to the initial shift of0.79±0.51mm.In the second experiment, we continue to use two cases of brain tumor patients in the clinical data to estimate the a accuracy of our system. According to our experiments, the average shift error after craniotomy was reduced from4.97±3.53mm to1.00±0.63mm and the average prediction rate is72.27±20.37%.According to this two experiments, the system was convenient and fast to operate and showed encouraging results and presented a promising pathway to compensating for brain deformation during neurosurgery.