Ancient pelvis reconstruction from collapsed component bones using statistical shape models

Ken’ichi Morooka, Ryota Matsubara, Shoko Miyauchi, Takaichi Fukuda, Takeshi Sugii, Ryo Kurazume

Research output: Contribution to journalArticlepeer-review


We propose a new method for recovering the pelvis of an ancient skeleton from its three component bones with collapsed surfaces. The proposed method uses four types of statistical shape models (SSMs) for the bones. The SSM for each bone describes the mean shape and shape variations of a class of bones. The SSMs for the three component bones are employed to restore the shapes of the component bones. The SSM for the whole pelvis provides the natural anatomical shape of the pelvis and the spatial relationship among the sacrum and the hip bones. Therefore, the three component bones are aligned by using the SSM for the pelvis. The experimental results show that our method achieves reliable reconstruction of the ancient pelvis shape despite having collapsed surfaces in its component bones.

Original languageEnglish
Pages (from-to)59-69
Number of pages11
JournalMachine Vision and Applications
Issue number1
Publication statusPublished - Feb 13 2019
Externally publishedYes


  • Ancient bone
  • Bone shape recovery
  • Statistical shape model

ASJC Scopus subject areas

  • Software
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Computer Science Applications


Dive into the research topics of 'Ancient pelvis reconstruction from collapsed component bones using statistical shape models'. Together they form a unique fingerprint.

Cite this