Feasibility of Individualized Airway Surgery in Horses
This thesis describes a progressive series of studies that was conducted to investigate the use of benchtop and computational models in the investigation of multiple upper airway surgeries for recurrent laryngeal neuropathy (RLN) in horses. Overall, the objective was to build the knowledge of laryngeal conformation and fluid mechanics with various surgical procedures for application in patients as an evaluation for the surgery of best outcome and as a step toward patient-specific diagnosis and treatment of upper airway disorders. The first study built upon a previously reported vacuum-based box setup that was used to compare RLN to four different surgical procedures within twenty-eight different equine larynges. Each larynx underwent these procedures in order and inhalation was simulated while measuring resistance to airflow (translaryngeal impedance). Two of the procedures, the combined laryngoplasty, ipsilateral ventriculocordectomy with arytenoid corniculectomy (LLPCOR) and laryngoplasty with ipsilateral ventriculocordectomy (LLP) were found to be significantly different from the RLN, arytenoid corniculectomy (COR) and partial arytenoidectomy (PA). The adapter used to mount the larynges was found to have a significant effect for the RLN, LLP and LLPCOR procedures. There was also a residual intraclass correlation of 27.6% in the final statistical model from individual laryngeal differences which were observed during the study. The variation of laryngeal features observed during the first study led to questions about the interaction between these geometries and airflow development. To capture the three-dimensional geometry effectively, similar methodology was repeated with concurrent computed tomography (CT) scans. These scans were analyzed focusing on cross-sectional area and changes along the airstream. Each procedural run was analyzed and used to simulate a pipe constriction. Entrance and exit conformations were modeled with respect to the ratio of the inlet cross-sectional area (CSA), constriction CSA and the divergent CSA downstream. The entrance characteristics were found to be significant; specifically, the angle of constriction and the ratio of the larger and smaller areas had a significant effect on laryngeal impedance. A frictional coefficient was measured as a function of energy lost by air passing through the constricted area and was found to be significant. This confirmed the importance of detail in surgically addressing disease affecting the laryngeal entrance. To provide a more thorough analysis of geometry and flow application of computational fluid dynamics (CFD) analysis was next reasonable step. The next study consisted of CFD analysis of the CT scans to determine the accuracy of CFD in reflecting the findings of the vacuum box airflow model. CFD provides a three-dimensional analysis of flow through complex geometries but also reduces the expense and intensive labor of complex flow experiments. Given these potential applications, this study reported the use of CFD to predict the procedure with the lowest impedance for each larynx. CFD results were compared to the measured values. Additionally, qualitative characteristics of the flow within the anatomical paradigm were examined. The CFD models corroborated the procedure of lowest impedance for 7 out of 10 of the larynges; 2 larynges had 2 procedures that were very close in impedance and the last larynx had unique collapse characteristics that may explain the lack of agreement. The measured pressure and impedance values showed a linear trend compared to the calculated values with measured impedance about 0.7 times that of the calculated (CFD) values. Qualitatively, areas of negative pressure and high velocity were noted in the higher impedance procedures and around tissue irregularities. While the CFD model was reasonably successful for the laryngeal study, demonstration of use in a more realistic equine patient application is needed. The final study took an additional step toward the equine patient by incorporating an entire head with measured translaryngeal impedance similar to the previous studies. A cadaver head was used and RLN, LLP, LLPCOR, COR and PA were simulated and subjected to negative airflow. The impedance values measured during this study were higher than expected, but the computational model reported values that were similar to the previous literature. The observed flow characteristics showed some differences to previous studies but the CFD model clarified these differences by highlighting the differences in three-dimensional geometry between the heads used in each study. The PA was the lowest impedance procedure both as measured and as calculated. Thus, CFD continued to demonstrate a predictive capability when it comes to determining the procedure of lowest impedance for the whole equine upper airway. Although there have been a large number of biomechanical models investigating the equine upper airway, they have not kept up with the technological advancements in human respiratory mechanics and CFD. CFD consistently confirmed the procedure of lowest impedance while incorporating individual patient geometry and can be performed much more efficiently than when the first equine application was reported over a decade ago. While more studies are needed, these models unquestionably provide a foundation for individual patient analysis in the future.
equine, veterinary surgery, recurrent laryngeal neuropathy, equine airway, respiratory mechanics, computational fluid dynamics, laryngoplasty, tie back, partial arytenoidectomy, arytenoid corniculectomy, equine upper airway
Doctor of Philosophy (Ph.D.)
Large Animal Clinical Sciences
Large Animal Clinical Sciences