Vibration-based elastic parameter identification of the diploë and cortical tables in dry cranial bones.

Abstract

Various human skull models feature a layered cranial structure composed of homogeneous cortical tables and the inner diploë. However, there is a lack of fundamental validation work of such three-layer cranial bone models by combining high-fidelity computational modeling and rigorous experiments. Here, non-contact vibration experiments are conducted on an assortment of dry bone segments from the largest cranial bone regions (parietal, frontal, occipital, and temporal) to estimate the first handful of modal frequencies and damping ratios, as well as mode shapes, in the audio frequency regime. Numerical models that consider the cortical tables and the diploë as domains with separate isotropic material properties are constructed for each bone segment using a routine that identifies the cortical table-diploë boundaries from micro-computed tomography scan images, and reconstructs a three-dimensional geometry layer by layer. The material properties for cortical tables and diploë are obtained using a Hounsfield Unit-based mass density calculation combined with a parameter identification scheme for Young's modulus estimation. With the identified parameters, the average error between experimental and numerical modal frequencies is 1.3% and the modal assurance criterion values for most modes are above 0.90, indicating that the layered model is suitable for predicting the vibrational behavior of cranial bone. The proposed layered modeling and identified elastic parameters are also useful to support computational modeling of cranial guided waves and mode conversion in medical ultrasound. Additionally, the diploë elastic properties are rarely reported in the literature, making this work a fundamental characterization effort that can guide in the selection of material properties for human head models that consider layered cranial bone.