The paper develops an AI-based segmentation approach integrating Bayesian level set with neural network to extract brain tissue and tumor. To obtain the neuron's weights in the neural network, we use the proposed Bayesian level set to segment simulated brain dataset and train the neural network. The neural network is then used to continuously extract the GM, WM, CSF, and tumor from a series of brain images. To better visualize the brain image, the extracted brain tissue and tumor are rendered by accelerated ray casting. Moreover, to accurately diagnose the brain tumor and understand its anatomical structure, the obtained surface area and volume are calculated by volume and surface integration.