A Novel Method for Classification and Characterization of Urothelium Cell Culture Exposed to the Different PPARg Agonists

The main purpose of the thesis is to classify and characterize urothelium cell cultures under PPARg activator/inhibitor. In this project, the raw data are obtained from videos of three different cell cultures: TZ (PPARg activator), T0070907 (PPARg inhibitor) and a control culture. A cell tracking program based on the OpenCV computer vision library is applied to the videos to generate a dataset consisting of x,y coordinates of the tracked cells. The numerical computing environment MATLAB® is subsequently used to filter the data and extract features, which were applied to machine learning algorithms to classify the cell cultures. Results obtained indicate that the TZ/T0070907 addition can cause a change in the average behavior of cells, such as the number of cells in the culture, the speed of cells and the average clump size of cells. The work also demonstrates that there is a difference in single cell behavior among different cultures. In summary, it is proposed that the approach described in this project provides a potential way of analyzing the average behavior of cells in different cultures.