An Intelligent Glioma Surgery-Assisted Navigation System Incorporating 3D-ResUNet and Mixed Reality

In this study, we propose an intelligent navigation system for glioma surgery that combines 3D-ResUNet and mixed reality technologies. The aim of this system is to enhance the accuracy and safety of glioma surgery. Automatic segmentation is achieved through the use of multimodal MRI data and the 3D-ResUNet model. Additionally, 3D reconstruction is performed using VTK to enable hierarchical visualization of brain tissues and tumor regions. By utilizing the Vuforia platform and HoloLens 2, the system is able to align and display virtual and real environments in real-time, providing surgeons with efficient surgical navigation support. Experimental results demonstrate that the system exhibits excellent segmentation performance, as indicated by the Dice coefficient and Jaccard index. It is anticipated that this system will be adopted and utilized in clinical surgery.