Automated Assessment of Small Bowel Motility Function Based on Feature Points Tracking

In this paper, we propose an automated method for assessing small bowel contraction movement with Cine-MRI based on automatic feature points tracking. The proposed method comprises two steps. First, in the initial frame, the user selects two arbitrary points on the boundary of the small bowel. The distance between the two points represents the size of the small bowl. Second, each of the two points is automatically tracked by the use of Kanade-Lucas-Tomasi (KLT) tracker method in the temporal sequence images. The contraction movement of the small intestine can be automatically analyzed by tracking these two points. Compared with the conventional method based on small bowel segmentation, The KLT-tracker method allows us to analyze the small intestinal contraction movement faster and establish a more objective analysis method.