Automatic Reconstruction of 3D Vascular Model for Mixed Reality

With the rapid developing of mixed reality technology in the medical field, and the using of mixed reality devices has attracted more and more people's attention, especially, researching on medical 3D model reconstruction technology. In this paper, a mixed reality automatic reconstruction method of 3D model is proposed for visualization, to achieve only once interaction needed in the reconstruction process. The method is to obtain the target tissue contour from each image, then to establish an ordered target tissue contour dataset, and finally to use the marching cube algorithm to reconstruct the contour dataset into a 3D model. The method was validated in CTA image data. The experiment has realized the automatic reconstruction of the neck vessel 3D model, and outputted in STL format, the 3D model has displayed in Hololens glasses to realize visualization. The experimental results show that the 3D model reconstructed and outputted by the method is effective. Reconstruction process of this approach has the characteristics of high reconstruction speed and only once manual interaction confirmation. And reconstruction result has a good effect in HoloLens glasses.