Contributions to Statistical Image Analysis for High Content Screening.
dc.contributor.author | Liu, Fangyi | en_US |
dc.date.accessioned | 2012-06-15T17:30:27Z | |
dc.date.available | NO_RESTRICTION | en_US |
dc.date.available | 2012-06-15T17:30:27Z | |
dc.date.issued | 2012 | en_US |
dc.date.submitted | en_US | |
dc.identifier.uri | https://hdl.handle.net/2027.42/91460 | |
dc.description.abstract | Images of cells incubated with fluorescent small molecule probes can be used to infer where the compounds distribute within cells. Identifying the spatial pattern of compound localization within each cell is very important problem for which adequate statistical methods do not yet exist. First, we asked whether a classifier for subcellular localization categories can be developed based on a training set of manually classified cells. Due to challenges of the images such as uneven field illumination, low resolution, high noise, variation in intensity and contrast, and cell to cell variability in probe distributions, we constructed texture features for contrast quantiles conditioning on intensities, and classifying on artificial cells with same marginal distribution but different conditional distribution supported that this conditioning approach is beneficial to distinguish different localization distributions. Using these conditional features, we obtained satisfactory performance in image classification, and performed to dimension reduction and data visualization. As high content images are subject to several major forms of artifacts, we are interested in the implications of measurement errors and artifacts on our ability to draw scientifically meaningful conclusions from high content images. Specifically, we considered three forms of artifacts: saturation, blurring and additive noise. For each type of artifacts, we artificially introduced larger amount, and aimed to understand the bias by `Simulation Extrapolation' (SIMEX) method, applied to the measurement errors for pairwise centroid distances, the degree of eccentricity in the class-specific distributions, and the angles between the dominant axes of variability for different categories. Finally, we briefly considered the analysis of time-point images. Small molecule studies will be more focused. Specifically, we consider the evolving patterns of subcellular staining from the moment that a compound is introduced into the cell culture medium, to the point that steady state distribution is reached. We construct the degree to which the subcellular staining pattern is concentrated in or near the nucleus as the features of timecourse data set, and aim to determine whether different compounds accumulate in different regions at different times, as characterized in terms of their position in the cell relative to the nucleus. | en_US |
dc.language.iso | en_US | en_US |
dc.subject | High Content Screening | en_US |
dc.title | Contributions to Statistical Image Analysis for High Content Screening. | en_US |
dc.type | Thesis | en_US |
dc.description.thesisdegreename | PhD | en_US |
dc.description.thesisdegreediscipline | Statistics | en_US |
dc.description.thesisdegreegrantor | University of Michigan, Horace H. Rackham School of Graduate Studies | en_US |
dc.contributor.committeemember | Shedden, Kerby A. | en_US |
dc.contributor.committeemember | Levina, Elizaveta | en_US |
dc.contributor.committeemember | Michailidis, George | en_US |
dc.contributor.committeemember | Rosania, Gustavo | en_US |
dc.subject.hlbsecondlevel | Statistics and Numeric Data | en_US |
dc.subject.hlbtoplevel | Science | en_US |
dc.description.bitstreamurl | http://deepblue.lib.umich.edu/bitstream/2027.42/91460/1/liufy_1.pdf | |
dc.owningcollname | Dissertations and Theses (Ph.D. and Master's) |
Files in this item
Remediation of Harmful Language
The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.
Accessibility
If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.