Age estimation of living individuals and the analysis of the phenomenon of long bone compression after epiphyseal closure based on non-invasive MRI image sequences of the knee.

对在世个人的年龄估计，特别是青少年和成年人之间的区别，是庇护法和刑法的重要组成部分。因此，使用了身体检查、手部X光检查以及牙科检查。由于辐射防护，射线照相只能在刑法的情况下使用。除了这一限制之外，大多数已发表的方法在应用于任何个人的情况下都应该被评为批判性的。这是因为大多数年龄估计方法都是基于对群体的统计分析，这些群体只涵盖与社会地位或出身有关的人口的一部分。非侵入性成像技术，如磁共振成像(MRI)，使评估骨痂生长的骨化过程成为可能。在这里，生长过程的等级与个体年龄之间的相关性被计算并用于实际年龄估计。但是，在大多数情况下，人工评估很耗时，评分也不清楚。这可能会导致依赖于用户的结果和错误分类。此外，使用图像分析方法进行自动和可靠的年龄估计，从而能够评估更多的数据集，目前还没有发表。另一种生物现象是长骨在完成生长过程后长度减少。在过去的几年里，这种收缩被附带地描述了一下，但没有测量纵向长度。2012年，第一项使用公里计进行纵向测量的研究发表。首次报道了胫骨近端骨骺完全闭合后的萎缩。本课题的目的是基于MRI图像序列对膝关节骨化程度进行评价，以建立一种新的年龄估计方法。应该开发一种新的方法，能够使用3D MRI数据集对骨化程度进行自动和稳健的分类。在分析大量的数据集时，应该研究这一程序在区分青少年和成人方面的适用性。此外，还应防止长骨受压的现象。因此，应在3年内多次获取和分析每个受试者的3D MRI图像数据、纵向测量和特征参数(如体重和身高)的新数据库。然后，根据MRI图像数据，使用基于计算机的方法计算胫骨的长度。之后，应将这些结果与公里计值进行比较，以证明基于计算机的方法的质量。此外，应该评估非线性配准以定位长骨萎缩的区域。最后，应该调查哪些提取的图像参数与法医年龄估计(16岁和18岁)相关，以及识别大多数年龄的精确度有多高。

Age estimation of living individuals especially the differentiation between adolescent and adult is an important part of asylum and criminal law. Therefore examinations of the body, radiography of the hand as well as odontology inspections were used. Due to radiation protection radiography can only use in case of criminal law. Beside this constraint most of the published methods should be rated critically in case of applying them to any individual. This is due to the fact that most of the age estimation methods are based on the statistical analysis of groups that covers only a subset of the population related to social status or origin. Non-invasive imaging techniques, like the magnetic resonance imaging (MRI), enable the assessment of the ossification process of epiphyses growth. Here, the correlation between the grade of the growth process and the age of the individuals is calculated and used for actual age-estimation. But, in most cases the manual assessment is time-consuming and the grading is not clear. This can lead to user dependent results and misclassification. Also the use of image analysis methods for automatic and robust age estimation enabling the evaluation of a greater number of datasets is not published yet. Another biological phenomenon is the length reduction of long bones after the completed growth process. In the last years this shrinkage was incidentally described but without measuring the longitudinal length. In 2012 the first study was published using a Knemometer for longitudinal measuring. Here, the shrinkage after complete closing of the epiphysis of the proximal tibia was documented for the first time. The aim of this project is the evaluation of the epiphyses ossification in the knee based on MRI image sequences to establish a new method for age estimation. A new method should be developed which enables an automatic and robust classification of the grade of ossification using 3D MRI datasets. The applicability of this procedure should be investigated for the differentiation between adolescent and adult analyzing a large collective of datasets. Additionally, the phenomenon of long bone compression should be proofed. Therefore, a new data base including 3D MRI image data, longitudinal measurement and characteristic parameters (e.g. weight and height) for each subject should be acquired and analyzed several times over a period of 3 years. Then, the length of the tibia should be calculated using computer based methods on MRI image data. Afterwards these results should be compared to the Knemometer values proving the quality of the computer based methods. Furthermore, non-linear registration should be evaluated to localize the area of long bone shrinkage. Finally, it should be investigated which of the extracted image parameters are relevant for forensic age estimation (16 and 18) and how precise the majority age can be identified.