Automatic Ribs Segmentation and Counting From Mouse X-ray Images

The automatic detection and quantification of skeletal structures has a variety of applications for biological research. This paper proposes an automatic solution for rib segmentation and counting based on structural properties of ribs in mouse X-ray images. The solution consists of five stages, including alignment, cropping the region of interest, image enhancement, rib segmentation, filtering of non-rib objects and rib counting. Experimental results on a set of 100 X-ray images showed consistent good performance of the algorithm compared with a manual annotation by domain experts, with an overall accuracy of about 88%.