Title Information
Type of Resource (primo): dissertations
Name: Personal
Name: Personal
Name: Personal
Name: Personal
Name: Personal
Name: Personal
Name: Corporate
Origin Information
Physical Description
Note: thesis: Thesis (Ph. D.)--Brown University, 2022
Genre (aat): theses
Abstract: Optical coherence tomography (OCT) is becoming increasingly popular for neuroscientific study, but it remains challenging to objectively quantify tissue and vascular properties from 3D or 4D OCT data. For tissue imaging, OCT provides structural imaging capability, which can be utilized for the determination of the attenuation coefficient. However, the accuracy of the attenuation coefficient is dependent on the precise characterization of the confocal effect and the position of the focal plane within the sample, which may be tilted or curved. This thesis presents a method to accurately determine the 2D position of the focal plane within the sample, which is also applicable to high-magnification data. We demonstrate applicability to real biological data of glioblastoma acquired in vivo in a murine model. For vascular imaging, OCT angiography can be performed to visualize perfused vessels, and capillary red blood cell (RBC) flux can also be quantified by acquiring time-series data at each 3D position. However, it remains challenging to quantify angioarchitectural properties from 3D angiograms, mainly due to projection artifacts or tails underneath vessels caused by multiple-scattering, as well as the relatively low signal-to-noise ratio compared to fluorescence-based imaging modalities. Challenges also exist in quantifying RBC flux which typically relies on peak-counting of transient peaks in intensity associated with RBC passages, and is dependent on user parameterization. To overcome these challenges, we leverage deep learning to enhance, segment and connect vessels in OCT angiograms, as well as to predict RBC flux from time-series data. These tools enable the extraction of the underlying 3D vascular graph and the subsequent quantification of vascular properties, as well as the estimation of RBC flux across an entire network of hundreds of capillaries. Finally, we leveraged these tools to track vascular changes in the same set of animals over a period of 7 months. These integrated methods enabled us to simultaneously monitor 24 distinct vascular properties spanning morphology, topology, and function of the microvasculature across all scales: large pial vessels, penetrating vessels, and capillaries. We demonstrate the efficacy of this approach in normal mice, and differential onset of vascular changes in mice that model mechanisms of Alzheimer’s disease. This new approach will allow comprehensive and longitudinal study of a broad range of progressive vascular diseases, and normal aging, in key model systems.
Subject (fast) (authorityURI="http://id.worldcat.org/fast", valueURI="http://id.worldcat.org/fast/00851361")
Subject (fast) (authorityURI="http://id.worldcat.org/fast", valueURI="http://id.worldcat.org/fast/00806532")
Subject (fast) (authorityURI="http://id.worldcat.org/fast", valueURI="http://id.worldcat.org/fast/01739425")
Subject
Language
Record Information