利用脑移位模型预测磁共振功能影像术中移位的初步研究

Object： In order to solve the brain shift in navigation, we used mathematical models based on thin-plate splines to adjust the navigative errors of DTI or BOLD . Methods: The techniques of DTI and BOLD were used to localize the pyramidal tracts and functional cortex in 19 patients with brain tumors involving pyramidal tracts in high-field MRI and 21 patients with tumors involving motor or linguistic cortex. In this research, pre-operative iMRI and intra-operative iMRI were used respectively for pre-operative and intra-operative data fields, and non-rigid calculation based on thin-plate splines was applied for predicting the tendency of brain shift. When this calculated tendency was integrated into DTI or BOLD images, predictive DTI or BOLD images after brain shift were acquired. Then we integrated these DTI or BOLD images into intra-operative iMRI images so as to visually learn the location and relative relationship of the brain tissue, residual tumor and pyramidal tracts or functional cortex after brain shift .Results: The imaging and integration were of success in total 40 patients. We also successfully acquired the prediction of DTI and BOLD images shift. Conclusions: The mathematical model of non-rigid calculation method based on thin-plate splines was effective in predicting the shift of DTI or BOLD images.