For cancerous tumors in vital internal organs, minimally invasive laser surgery may be a desirable choice for cancer treatment due to its precise control and compatibility with most of the imaging modalities such as MRI (magnetic resonance imaging). However, the complexity of tumor composition and tissue response to a thermal dose demands real time optimization and control. In the previous work, we have developed a quite general computational framework that is capable of processing MRI anatomical data, providing pretreatment surgical protocol, and controlling tissue damage based on in vivo MRTI (magnetic resonance thermal imaging) data. In this paper, we describe computational techniques that are involved in real time optimization and control for laser surgical protocols of cancer treatment.