3D Reconstruction Method Research Of The Brain Structure Based On The Talairach Atlas

In the recent years, research on human brain has become one of the research hotspots. 3D visualization of Brain mapping plays a decisive role in the subsequent surgery plan, virtual surgery operation and teaching applications.Talairach-Tournoux atlases (also known as Talairach atlas) have been one of the extensive reference brain atlases by clinical doctors. The atlas were from an cadavers specimens of brain, outlined organizational structure by cutting the brain specimens, and described brain anatomy information with color graphics. The interval between two slices is relatively large spacing from 2 mm to 5mm. For this feature of Talairach atlas, more commonly used three-dimensional reconstruction techniques (such as Marching cube algorithm) very difficult to achieve ideal rebuilding effect. This paper presents a shape-based interpolation method to reconstruct organizational structure of the brain atlas.Firstly, extract the contours of the designated organizational structure in Talairach atlas, use polygon approximation to approximate contours, when the two contours reached the maximum similarity between the matching effects, use characteristics of shape interpolation to interpolate needs contours. Secondly, all contours are grouped into simple contours group (one-to-one) and the complex contours group (one-to-many or many-to-many), for the simple contours group, using NURBS-based method to reconstruct 3D surface; for complex contours group, using distance map-based method to rebuild 3D surface. Finally, achieve the smooth connecting of the 3D surface, and form the 3D surface of this organization. In TT atlas, because of the one-to-many contours situation and the many-to-many situation, after finding the right relationship between the upper and lower contours, what must solve the problem is bifurcation contours, make complex contours group into simple contours group, and make reconstruction simplify. Experiments show that the method can be achieved very good reconstruction results not only for the normal tissues but also for the pathological changes tissues.This method is not only suit for the Talairach atlas' 3D visualization, but also for the CT\MRI segmentation image with large slice interval. The method has the in-depth theoretical and practical value for the clinical application of digital brain atlas.