Rectal wall structure delineation and broken layer recognition by multigradient field active contour

This paper presents a novel multigradient field active contour algorithm with an extended ability for multiple object delineation, which overcomes some limitations of ordinary active contour model (snakes). One of the aims is to apply this technique for multilayer boundary detection from ultrasound rectal wall image, which is important in colorectal clinical diagnosis for rectal tumor staging. The core part in the algorithm is the proposal of multigradient vector field concept, which is used to replace the image forces in the kinetic equation of the active contour. For some broken layers on the rectal wall from the penetration by tumor, a modified local cost function approach is proposed to recognize the broken segments from the current obtained boundaries.