Template-based Shape Modeling and Anatomical Region Mapping of Bone


In the bone shape morphology, the morphological difference of the bones across subjects is quantified by the geometric measures, such as the curvature of the articular surface and the relative bone orientation in joints, defined with the anatomical landmarks and regions on the bone surface. However, the landmark and region determination on individual cases is a difficult and time-consuming task, because of the various size and shape of the bones and operator’s errors.

To address this problem, we introduce our template-based shape modeling framework to a posture and gait analysis on foot bones and joints to automatically determine region landmarks (e.g. articular surfaces) of bones on their surfaces. We proposed a landmark and region mapping method based on a non-rigid template-to-image registration, an automated method for extracting the joint space by applying graph cut on Markov random field model, and reliable measurement of 3D foot bone angles.

Articular surface and landmarks, which are automatically identified for the calcanei of 3 subjects

Template model of talus and navicular with the articular surfaces encoded

Joint space extraction results. (a) Specified ROI (blue) and extracted joint space (red) between the bones using the proposed method. (b) Separated bone segments shown in 3D.

3D visualization of the segmented joint space (shown in red) using the proposed method. From left to right, calcaneocuboid joint, talonavicular joint, and talocalcaneal joint.

Extracted sole (red) and two feature points (yellow)

(Left) Centers of hallux and 4th toe. (Right) Refined position of hallux and 4th toe (the tips of the arrows).

(Left) Directions of the 1st and the 2nd axes with the extracted feature points. (Right) The 3rd axis is a distal to proximal direction, which is a cross product of the 1st and the 2nd axes.


Taeho Kim (kdhtheo at kaist.ac.kr)