Segmentation of the Prostate Gland in Images Using Prior Knowledge and Level Set Method

The treatment of prostate cancer is a great challenge, with very precise indications due to the high morbidity and mortality rates. Magnetic resonance imaging (MRI) is used to monitor the disease, providing a noninvasive clinical diagnosis to the patient. In this sense, this work proposes a segmentation of magnetic resonance imaging in models to identify a prostatic area, delimiting its area. To do so, a new approach to segmenting base images into a priori knowledge of the form and method Level Set will be developed. The performance of the proposed segmentation algorithm was evaluated with respect to the manual segmentation ground truth, active shape model and the level set method without priori knowledge. The experimental results obtained show that the inclusion of a priori information and the constraint imposed by the shape model contribute to greater segmentation accuracy than other methods.