A well-established method for treating lung cancer is curative-intent radiation therapy (RT). The most significant challenge for RT is to accurately target the lesion volume while avoiding the irradiation of surrounding healthy tissue. Currently at the Juravinski Cancer Centre (JCC), treatment plans for lung cancer patients are completed using fluorodeoxyglucose positron emission tomography (FDG-PET) and four-dimensional computed tomography (4DCT) images. There is no clear protocol, however, to compensate for respiratory motion in PET images and it is not known how lesion volumes generated from PET reflect the true volume. This project evaluated methods to optimize the use of PET images in the radiation treatment planning workflow and quantify the effects of respiratory motion. First, a 4D XCAT digital phantom was used to quantify respiratory motion and its effects on lesion displacement. A CTN physical phantom and 3D-printed irregularly shaped lesion were imaged to determine the accuracy of the PET EDGE automated segmentation algorithm (ASA). Lastly, rigid and deformable image registration techniques were used to propagate the diagnostic PET scan of the irregular lesion to the 4D planning CT. PET EDGE was used to generate target volumes which were then compared to internal target volumes (ITVs) generated from manual contouring of the 4DCT image alone. We found that lesion displacement due to respiratory motion can be adequately modeled using a moving platform set to oscillate 1 cm and 2 cm for normal and deep breathing, respectively. Optimal target delineation was found when diagnostic PET was propagated to the planning CT using rigid image registration for lesions that experienced 1 cm of oscillatory motion during imaging. In contrast, PET EDGE would overestimate volumes in static cases and underestimate volumes in instances of 2 cm dynamic motion meant to simulate deep breathing. / Thesis / Master of Science (MSc)

Background: Before making any treatment decisions for patients with non-small cell lung cancer (NSCLC), it is crucial to determine whether the cancer has spread to the mediastinal lymph nodes (LNs). The preferred method for mediastinal staging is Endobronchial Ultrasound Transbronchial Needle Aspiration (EBUS-TBNA). However, EBUS-TBNA has been reported to generate inconclusive results as much as 40% of the time. Since this jeopardizes good patient care, there is near-universal consensus on the need to develop and study a novel method for LN staging. Recent research has shown that AI and deep learning are used to accurately interpret images with comparisons to clinicians in radiology, pathology, and cardiology. Additionally, EBUS-Elastography is a novel modality which could be used as an adjunct to EBUS-TBNA for LN staging. This technology uses impedance ultrasonography to measure tissue stiffness. Methods: There are three parts to this thesis. The first part involved the training, validating, and testing NeuralSeg, a deep neural network, to predict LN malignancy based on B-mode EBUS-TBNA images. The second part of this thesis involves EBUS-Elastography, defining the blue colour threshold and the optimal SAR cut-off value based on the blue threshold that most accurately distinguished benign and malignant LN. Finally, this thesis's third part involves validating part II's findings. Results: Part I resulted in an overall accuracy of 80.63% (76.93% to 83.97%), a sensitivity of 43.23% (35.30% to 51.41%), a specificity of 96.91% (94.54% to 98.45%), a positive predictive value of 85.90% (76.81% to 91.80%), a negative predictive value of 79.68% (77.34% to 81.83%), and an AUC of 0.701 (0.646 to 0.755). Part II Level 60 was chosen as the blue threshold with an AUC of 0.89 (95% CI: 0.77-1.00), and the optimal SAR cut off was found to be 0.4959 with a sensitivity of 92.30% (95% CI: 62.10% to 99.60%) and a specificity of 76.50% (95% CI: 49.80% to 92.20%). Using the blue threshold and SAR cut-off, the results of part III resulted in an overall accuracy of 70.59% (95% (CI) 63.50% to 77.01%), the sensitivity of 43.04% (CI: 31.94% to 54.67%), and a specificity of 90.74% (CI: 83.63% to 95.47%). Conclusion: It was observed that both AI and AI-powered EBUS-Elastography achieved high specificities on larger sample sizes, indicative that these methods may be helpful in identifying LN malignancy. However, due to the novelty of these technologies, more extensive multi-centre studies must be conducted before these processes can be standardized. / Thesis / Master of Health Sciences (MSc) / Non-Small Cell Lung Cancer (NSCLC) treatment decisions are made using vital information by performing biopsies to collect tissue from the lymph nodes near the lungs. The current method is called Endobronchial Ultrasound Transbronchial Needle Aspiration (EBUS-TBNA), which involves a scope with a fine needle attached to it. This scope is led down the airway and guided by ultrasound to obtain the tissue needed to determine whether the lymph nodes have cancerous tissue. If the lymph nodes contain cancerous tissue, the patient may need chemotherapy; however, lung surgery may be the best treatment option if they do not. Many factors impact how successfully these tissue samples can be obtained, such as the skill and experience of the surgeon. These factors often lead to inconclusive results, making it difficult to make correct treatment decisions. Novel technologies such as Artificial Intelligence and Elastography are being used to diagnose lung cancer by interpreting images and providing information on tissue stiffness. We trained an Artificial Intelligence program to predict malignancy based on EBUS-TBNA images. Additionally, we trained the AI program to analyze Elastography images to aid us in understanding the relationship between the colour pattern of the elastography images and cancerous tissue. This thesis assesses how these novel technologies contribute to lung cancer diagnosis.

The laminar flow in the small bronchial tubes is quite complex due to the presence of vortex-dominated, secondary flows. Contributing to this complexity are the geometrical characteristics of the bronchial tubes that include asymmetric and nonplanar branching. These secondary flow fields play a crucial role in particle deposition; however, the actual mechanisms that determine the particle distributions are not fully understood. The research reported here increases understanding of this phenomenon by studying flow structure/ particle interaction in the small bronchial tubes for steady and unsteady respiratory conditions. Specifically, the effects of simultaneous nonplanar and asymmetric branching were investigated. The nonplanar model was generated by applying a 90◦ out-of-plane rotation to the third-generation branches. Steady-state inspiratory flows for a Reynolds number of 1,000 and unsteady periodic flows with a 30-respiration-per-minute breathing frequency were simulated in three-generation, asymmetric, planar and nonplanar models. The asymmetry and nonplanarity produced asymmetric secondary flow patterns and unequal mass flow partitioning in the third-generation branches. Ten micron water droplet deposition in the nonplanar model was found to be significantly different from the planar model, demonstrating the impact of simultaneous nonplanar and asymmetric branching. The unsteady nature of the flow also affected particle deposition. Particles released at the same instantaneous inflow conditions during off-peak inhalation conditions, generated significantly different particle deposition patterns. The differences were attributed to the high temporal variations of the fluid velocities at these off-peak times and history effects in the flows. It was also observed that the initial particle velocities had a significant impact on particle deposition. The study of flow structure and particle interaction was facilitated by the development of a novel visualization technique that employs finite-time Lyapunov exponents (FTLE). This research provides a better understanding of the fluid dynamics driving the particle deposition in the bronchial tubes.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.