{"title":"Synthesis of an individualized cranial atlas with dysmorphic shape","authors":"G. Christensen, A. Kane, J. Marsh, M. Vannier","doi":"10.1109/MMBIA.1996.534083","DOIUrl":null,"url":null,"abstract":"A new method for non-rigid registration of a normal infant CT head atlas with CT data of infants with abnormal skull shape is presented. An individualized atlas is synthesized by computing a volume transformation from the normal atlas to the target data set shape. This process begins rigidly by eliminating translation and rotation differences and proceeds non-rigidly to eliminate anatomical shape differences. Operator specified anatomical landmarks are used to find the initial rigid transformation. Non-rigid registration is achieved by constraining the transformation by the low frequency modes of vibration of a 3D linear-elastic solid while minimizing the squared intensity difference between atlas and target CT image volumes. Results are presented in which the CT atlas was transformed into the shape of several infants with various types of craniofacial deformities.","PeriodicalId":436387,"journal":{"name":"Proceedings of the Workshop on Mathematical Methods in Biomedical Image Analysis","volume":"58 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1996-06-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"22","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the Workshop on Mathematical Methods in Biomedical Image Analysis","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/MMBIA.1996.534083","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 22
Abstract
A new method for non-rigid registration of a normal infant CT head atlas with CT data of infants with abnormal skull shape is presented. An individualized atlas is synthesized by computing a volume transformation from the normal atlas to the target data set shape. This process begins rigidly by eliminating translation and rotation differences and proceeds non-rigidly to eliminate anatomical shape differences. Operator specified anatomical landmarks are used to find the initial rigid transformation. Non-rigid registration is achieved by constraining the transformation by the low frequency modes of vibration of a 3D linear-elastic solid while minimizing the squared intensity difference between atlas and target CT image volumes. Results are presented in which the CT atlas was transformed into the shape of several infants with various types of craniofacial deformities.