Tumor Detection Based on Bayesian Level Set Method and Improved Marching Cubes Algorithm

We propose an approach, integrating Bayesian level set method with improved marching cubes algorithm for brain tumor detection. First, we extend the level set method based on the Bayesian risk to three-dimensional segmentation. Then, the three-dimensional Bayesian level set method is used to segment solid three-dimensional targets (e.g., tissue, whole brain, or tumor) from serial slice of medical images. Finally, the improved marching cubes algorithm is used to continuously reconstruct the surface of targets. Since each step can definitely obtain an appropriate treatment by statistical tests, the tissue and tumor segmentation and surface reconstruction are expected to be satisfied