Raw Data, PC scores, and secondary variables

The skull base serves as structural support for the brain and an important conduit for cranial nerves and vessels. Paraclival and parasellar features are particularly complex, as these sites represent the union of multiple bony elements with many neurovascular components. Although morphometric descriptions of these components are common, these methods have intrinsic limitations which make global assessment of these anatomical features and their relationships very challenging. Here we apply geometric morphometric (GM) techniques to address the problems associated with traditional morphometric strategies of evaluating skeletal and soft tissue components of the skull base. Our aims are to describe the geometry of relevant skull base features and assess the relationship between 3-D shape variables and more traditional craniometric variables. A Microscribe® i+ 3-D digitizer was used to register the location of 20 3-D skeletal and neural landmarks on the skull base of the middle and posterior cranial fossae on 80 anatomical body donors (mean age 83.0 ± 10.1 years). Linear measurements of neurocranial length, width, skull base length, and ulnar length (as a proxy for general body size) were collected with traditional methods. A principal component analysis (PCA) was applied to Procrustes aligned coordinates. The results of the PCA revealed 27 principal components (PCs). The first component accounted for 34.24% of the variation in the sample and was associated with changes in clival width with corresponding changes in sella turcica length. Shape changes associated with PC1 scores correlated with configuration size (centroid size) (R2=.532, p<.001). Principal component 2 (12.85%) was associated with isolated changes in clival width driven primarily by the jugular tubercles. Results related to PC3 (11.10%) showed variation in skull base flattening. While PC2 and PC3 results were not correlated with size, they were correlated with neurocranial length and measurements of cranial width. In summary, we were able to determine the primary axes of 3-D morphological variation in the skull base features we assessed. Skull base size emerged as a principal driver of skull base shape. Variable anatomy of the jugular tubercle and its relationships to the jugular foramen and hypoglossal canal represented another key component to shape variation and should be considered in surgical approaches to the posterior fossa. Finally, our results support the idea that a wide range of non-pathological skull base flattening may exist in the population, as this shape axis was a major contributor to the variation seen in the sample.