ABSTRACT
Prostate diseases, such as prostate cancer, prostatitis, and enlarged prostate, are common in males. In addition to its high incidence, prostate cancer also brings high mortality. According to the American Cancer Society, prostate cancer was the second leading cause of cancer death in American men in 2016, behind only lung and bronchus cancers. A set of works, including contour-based methods and deformable model-based methods, exploit the contours and shape information in prostate segmentation. Since deformable model was first introduced by Terzopoulos, it has been widely applied in many prostate segmentation works using contour and shape information. Random walker could be an effective and efficient algorithm to solve the labelling propagation problem in prostate segmentation. The classic classifiers, such as support vector machine and random forest, have been extensively studied in the last decades and proved favorable capacity of feature space partition, thus can also be applied in prostate segmentation works.
