Carneiro, JorgeSilva, Pedro Ângelo Pereira da2019-06-272019-06-2720172017-04http://hdl.handle.net/10362/73810"Digitization and robotization of laboratory equipment has recently contributed to the generation of high content of data and its metadata. While this seems like an advantage for science's celerity, the analysis of such data became the limiting step { a very narrow bottleneck. Such is the case for imaging data acquisition and its analysis. After collecting Gigabytes of images, researchers spend several orders of magnitude of more time to determine the regions of interest (ROIs) (e.g. cell) and to measure relevant attributes (e.g. mean uorescence intensity). This manual curation of data promotes another issue that is related with the reproducibility of the analysis, e.g., the same researcher will hardly select the exact same ROIs in the same data set. Furthermore, there is also the possibility of bias in the selection of which cells to use in the analysis by biased determination of the ROIs.(---)"engSea urchin Spermatozoaimage analysiscellsmorphodynamicaldoctoral thesis