Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article

Degree

Doctor of Philosophy

Program

Biomedical Engineering

Supervisor

Jenkyn, Thomas R.

Abstract

Concussion is a traumatic brain injury usually caused by a direct or indirect blow to the head that affects brain function. The maximum mechanical impedance of the brain tissue occurs at 450±50 Hz and may be affected by the skull resonant frequencies. After an impact to the head, vibration resonance of the skull damages the underlying cortex. The skull deforms and vibrates, like a bell for 3 to 5 milliseconds, bruising the cortex. Furthermore, the deceleration forces the frontal and temporal cortex against the skull, eliminating a layer of cerebrospinal fluid. When the skull vibrates, the force spreads directly to the cortex, with no layer of cerebrospinal fluid to reflect the wave or cushion its force. To date, there is few researches investigating the effect of transient vibration of the skull. Therefore, the overall goal of the proposed research is to gain better understanding of the role of transient vibration of the skull on concussion. This goal will be achieved by addressing three research objectives. First, a MRI skull and brain segmentation automatic technique is developed. Due to bones’ weak magnetic resonance signal, MRI scans struggle with differentiating bone tissue from other structures. One of the most important components for a successful segmentation is high-quality ground truth labels. Therefore, we introduce a deep learning framework for skull segmentation purpose where the ground truth labels are created from CT imaging using the standard tessellation language (STL). Furthermore, the brain region will be important for a future work, thus, we explore a new initialization concept of the convolutional neural network (CNN) by orthogonal moments to improve brain segmentation in MRI. Second, the creation of a novel 2D and 3D Automatic Method to Align the Facial Skeleton is introduced. An important aspect for further impact analysis is the ability to precisely simulate the same point of impact on multiple bone models. To perform this task, the skull must be precisely aligned in all anatomical planes. Therefore, we introduce a 2D/3D technique to align the facial skeleton that was initially developed for automatically calculating the craniofacial symmetry midline. In the 2D version, the entire concept of using cephalometric landmarks and manual image grid alignment to construct the training dataset was introduced. Then, this concept was extended to a 3D version where coronal and transverse planes are aligned using CNN approach. As the alignment in the sagittal plane is still undefined, a new alignment based on these techniques will be created to align the sagittal plane using Frankfort plane as a framework. Finally, the resonant frequencies of multiple skulls are assessed to determine how the skull resonant frequency vibrations propagate into the brain tissue. After applying material properties and mesh to the skull, modal analysis is performed to assess the skull natural frequencies. Finally, theories will be raised regarding the relation between the skull geometry, such as shape and thickness, and vibration with brain tissue injury, which may result in concussive injury.

Summary for Lay Audience

A concussion is a traumatic brain injury usually caused by a direct or indirect blow to the head that affects brain function. As the maximum mechanical impedance of the brain tissue occurs at 450±50 Hz, skull resonant frequencies may play an important role in the propagation of this vibration into the brain tissue. The overall goal of the proposed research is to gain a better understanding of the role of transient vibration of the skull on concussion. This goal will be achieved by addressing three research objectives: I) develop an automatic method to segment/extract skull and brain from magnetic resonance imaging (MRI), II) create a novel 2D and 3D automatic method to align the facial skeleton, and III) identify the skull resonant frequencies and raise the theory of how these vibrations may propagate into brain tissue. For objective 1, 58 MRI and their respective computed tomography (CT) scans were used to create a convolutional neural network framework for skull and brain segmentation in MRI. Moreover, an invariant moment kernel was introduced to improve the brain segmentation accuracy in MRI. For objective 2, a 2D and 3D technique for automatically calculating the craniofacial symmetry midline from head CT scans using deep learning techniques was used to precisely align the facial skeleton for future impact analysis. In objective 3, several skulls segmented were tested to identify their natural resonant frequencies. Those with a resonant frequency of 450±50 Hz were selected to improve understanding of how their shapes and thickness may help the vibration to propagate deeply in the brain tissue. The results from this study will improve our understanding of the role of transient vibration of the skull on concussion.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Download

Share

COinS