Abstract—Landmark correspondence is one of the key steps in Statistical Shape Model (SSM) building. In this paper, a non-rigid iterative closest point surfaces registration method is introduced to seek proper corresponded landmarks in the multi-organ surface meshes. Surfaces of four abdominal organs are used in the experiment to build SSM from landmarks corresponded by five different registration strategies. The proposed method of individually non-rigid registration of single organs shows the least errors measured by Hausdorff distance and the best model quality of generalization ability, specificity, and compactness.

Index Terms—multi-organ, statistical shape model, surface registration, landmark correspondence

Cite: Jiaqi Wu, Guangxu Li, and Tohru Kamiya, "Multi-organ Statistical Shape Model Building Using a Non-rigid ICP Based Surface Registration," Journal of Image and Graphics, Vol. 10, No. 3, pp. 95-101, September 2022.

Copyright © 2022 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.