Development Of Software For Neurosurgery Navigation System Based On FNIRS

The most important effect on neurosurgery operations is the location of target. At present, the main surgical navigation systems based on CT and MRI all have the problems which are hard to overcome: the displacement of brain tissue coursed by cerebrospinal fluid bleeding and cerebral swelling, and unable to recognize the tissue and target in real-time. As a new way of neurosurgical navigation, fNIRS technology can effectively solve the problems. It can get the best correlation factor between fNIRS and MRI, and then establish the associated mathematical model. Through this model, the gray-scale information of puncture track can be translated into optical parameters as prior knowledge. On the basis of these, the correction model which is the image information of puncture trajectory cone neighborhood is established to achieve real-time track error correction.In this paper, the key technologies of fNIRS assisted navigation included: the relevance of fNIRS and MRI data, three-dimensional visualization and surgical path planning and correction which are studied. First of all, through Phantom experiments show that the curve of reduced scattering coefficient changing by depth has great relativity with gray value on the same surgical puncture path; Secondly, in the data processing of MRI, an improved fast algorithm of ray casting volume rendering is presented which can produce the required quality images, at the same time speed up the volume rendering, and the semi-transparent processing and arbitrary profile cutting are achieved on the base of three-dimensional volume data; And then, the paper studies on the technique of path information extraction based on three-dimensional volume data which using the Hausdorff-distance curve curvature matching algorithm, determining the position of the target path and calculating coordinates slippage has achieved real-time path correction; Finally, the paper finished the software interface using VTK based on VC.The innovations of this study are as follows: (1) putting forward the idea of applying fNIRS technique in surgical navigation to solve the problem of target excursion, (2) improving the traditional ray casting algorithm and advancing a synthetical ray casting algorithm which speeded up the volume rendering under the premise of required quality images, (3) using the Hausdorff-distance algorithm to achieve matching of path information and guide real-time correction in surgery.