Software-based tools for biomedical imaging : from single cells to transplantation
Abstract/Contents
- Abstract
- Cancer is a group of serious diseases the world struggles to overcome, with over 14 million new cases and nearly 9 million deaths each year. Early detection greatly improves prognosis for nearly all types of cancer. In cases of later detection, treatments such as autologous stem cell transplants carry inherent risk of reintroducing cancer cells back into the patient, and current methods to combat this, such as cell sorting and purging, remain insufficient. Culturing a patient's harvested stem cells, monitoring their behavior, and assessing their safety and efficacy through non-invasive, label-free methods could greatly reduce the risk of graft failure and tumor formation after transplantation. Monitoring of the injection site post transplantation via implantable, miniaturized microscopes will allow us to follow these cells as they engraft and functionalize, serving as a second layer of safety for transplantation treatments, as a means toward early stage detection of cancers, or as a tool for studying early tumor formation. In this dissertation, I present four novel computational microscopy methods aimed at imaging and scrutinizing cells from single cell densities with bench-top sized optics to patient transplant sites with miniaturized optics. When culturing cells at low (single cell) to medium-high (nearly confluent) densities, automated cell segmentation and tracking is the ultimate goal, capable of quantifying communication, migration, morphology, and organization. However, accurate, automated, and easy-to-use cell segmentation remains a challenge, especially in cases of high cell densities, where discrete boundaries are not easily discernable. I have developed a fully automated segmentation algorithm that iteratively segments cells based on the observed distribution of optical cell volumes measured by quantitative phase microscopy. By fitting these distributions to Gamma probability density functions, the algorithm converges on volumetric thresholds that enable valid segmentation cuts. Since each threshold is determined from the observed data itself, virtually no input is needed from the user. The algorithm is demonstrated using six cell types displaying an extremely wide range of morphologies and confluency rates, demonstrating invariance to cell line, confluency rate, morphology, and time. As cultures become too confluent to identify the boundaries of individual cells, we must move to other techniques to analyze the available data. To that end, I present a machine learning technique called digital texture analysis for analyzing cell colonies. The process begins by quantifying image blocks using gray-level co-occurance matrices, then using said matrices to compute rotationally invariant Haralick textural features. Linear discriminant analysis reduces the dimensionality of feature space while preserving class discrimination. Support vector machines forms decision boundaries in eigenfeature space. The algorithm is tested on phase contrast timelapse data of human embryonic stem cells, showing correlation between loosely packed cells at the boundaries of colonies and differentiation status. Once the cell culture process is complete, we will use miniaturized laser scanning confocal microscopes to monitor the injection site after the cell culture process is complete. However, before moving directly to 3D in vivo imaging, I first present software-based methods for solving the three main issues with these microscopes in 2D: (1) slow frame rates, (2) phase control, and (3) small FOVs. To achieve video-rate imaging, I artifically boost the image update rate tenfold from 3 to 30 Hz, then vertically interpolate each sparse image in real-time to eliminate fixed pattern noise. To recover the resolution lost through spare sampling, and to greatly increase FOV, I mosaic consecutive images together. This is demonstrated by imaging fixed mouse brain tissues at varying update rates and compare the resulting mosaics. To control phase using software, I developed an algorithm that minimizes the high-frequency artifacts introduced through phase inaccuracies in the image mapping process. I validate this algorithm by imaging fluorescent beads and automatically maintaining phase control over the course of one hour in real-time. Using reconstructed image data as feedback for phase control eliminates the need for phase sensors and feedback controllers, enabling long-term imaging experiments without additional hardware. Mosaicing subsampled images results in video-rate imaging speeds, nearly fully recovered spatial resolution, and millimeter-scale fields of view. Finally, to enable cellular level 3D in vivo imaging of stem cell transplantation sites, I have helped developed an implantable dual-axis confocal microscope equipped with a 3D scanning system that consists of a 2D lateral scanner and a 1D vertical scanner. The microscope is optically characterized with transverse and axial resolutions of 2.8 um and 22 um, respectively. By implementing the first nonrepeating 3D Lissajous scanning patterns for an optical imager, volumes are sampled at rates between 0.5 Hz and 0.05 Hz, with fill-factors of 79% and 100%, respectively. Fluorescent bead phantoms and fluorescently labeled murine brain tissues are used to demonstrate high-resolution imaging in 3D.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2018; ©2018 |
| Publication date | 2018; 2018 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Loewke, Nathan Owens |
|---|---|
| Degree supervisor | Solgaard, Olav |
| Thesis advisor | Solgaard, Olav |
| Thesis advisor | Contag, Christopher H |
| Thesis advisor | Pauly, John (John M.) |
| Degree committee member | Contag, Christopher H |
| Degree committee member | Pauly, John (John M.) |
| Associated with | Stanford University, Department of Electrical Engineering. |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Nathan O. Loewke. |
|---|---|
| Note | Submitted to the Department of Electrical Engineering. |
| Thesis | Thesis Ph.D. Stanford University 2018. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2018 by Nathan Owens Loewke
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...