Research Experience and Mentoring

Background: Our lab is working on methods to segment and classify cell nuclei in 3D images of tissue samples. As a baseline for this, we have to develop an algorithm to identify cells in standard 2D tissue images using traditional fluorescence microscopy. The goal of this project is to train the candidate in acquiring, labeling, and imaging 2D tissue sections using a fluorescent microscope. They will then use the StarDist algorithm to create an automated tool to identify cell nuclei in images of organ cross-sections.

Research Plan: The candidate will be trained in the following: (1) basic histology, including how to cut tissue sections from archival paraffin blocks; (2) how to label tissue samples, including deparaffinization and fluorescent staining; (3) the basics of fluorescence microscopy and how to collect images using an inverted wide-field fluorescence microscopy; (4) how to use StarDist, a machine learning tool for cell nuclei segmentation; and (5) how to apply their trained StarDist algorithm to locate and classify all cells in their acquired microscope images.