Please use this identifier to cite or link to this item:
http://hdl.handle.net/11375/27802
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Zhang, Boyang | - |
dc.contributor.author | Abdul Majeed Abdul Wadood, Lyan | - |
dc.date.accessioned | 2022-09-13T14:02:33Z | - |
dc.date.available | 2022-09-13T14:02:33Z | - |
dc.date.issued | 2022 | - |
dc.identifier.uri | http://hdl.handle.net/11375/27802 | - |
dc.description.abstract | Three-dimensional (3D) tissue models, like organoids and spheroids, have been used to recapitulate many characteristics of human tissues. In particular, 3D tissue models can mimic key structural features of the native tissue. However, analyzing the morphology of 3D tissue models is limited by current image analysis tools that extract features like size, contours and texture which cannot fully describe the exhibited morphologies. Instead, deep learning techniques, which can identify objects from images directly can be used to make this type of analysis possible. In this work, we focus on applying deep learning to analyze the 3D tissue morphologies in brightfield images for the first time. Specifically, we developed and validated deep learning models to analyze important structures exhibited by lung spheroids or colon organoids. Deep-LUMEN is a custom deep learning model that can analyze the polarization of lung spheroids. We validated Deep-LUMEN by assessing how the extracellular matrix affects spheroid polarization and how the drug cyclosporin disrupts spheroid assembly. By analyzing the morphological features of lung spheroids, we found that cyclosporin can induce toxic effects at much lower concentrations than expected. This work also presents D-CryptO, a deep learning-based tool that can be used to analyze the structural maturity of colon organoids. D-CryptO analyzes the opacity and the presence of budding structures to assess tissue maturation. We validated D-CryptO by analyzing colon organoid morphology during prolonged culture and short-term perturbation with external stimuli. Additionally, we further applied it to assess organoid morphology following treatment with several chemotherapeutic drugs. Using D-CryptO, we gained insights into potential mechanisms of drug-induced toxicity. Together, these models demonstrate that deep learning is a viable technique to analyze 3D tissue morphology and it can be applied in a broad range of biological studies to gain useful insights into tissue physiology. | en_US |
dc.language.iso | en | en_US |
dc.title | Analyzing the Morphology of Three-Dimensional Tissue Models using Deep Learning | en_US |
dc.type | Thesis | en_US |
dc.contributor.department | Biomedical Engineering | en_US |
dc.description.degreetype | Thesis | en_US |
dc.description.degree | Master of Applied Science (MASc) | en_US |
Appears in Collections: | Open Access Dissertations and Theses |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
AbdulMajeedAbdulWadood_Lyan_finalsubmission2022June_M.A.Sc.pdf | 1.98 MB | Adobe PDF | View/Open |
Items in MacSphere are protected by copyright, with all rights reserved, unless otherwise indicated.