Ústav molekulární genetiky AV ČR, v. v. i.
The course concentrates on image data segmentation and cell tracking through the application of cutting-edge deep learning methods, with a specific emphasis on their practical applications.
This course focuses on image data segmentation and cell tracking using state-of-the-art deep learning methods like StarDist, Cellpose, Omnipose, and MitoSegNet. It shows preprocessing data by using AI methods like Noise2Void etc. It demonstrates how segmentation aids in analyzing image-based objects and how tracking applies segmentation to study cells dynamically over time. TrackMate, a plugin in Fiji, integrates StarDist and Cellpose for cell segmentation and tracking. The course also covers the Delta2 framework for tracking dense bacteria populations and explores ZeroCostDL4Mic, a cloud computing framework with advanced deep learning methods for microscopy tasks like segmentation, object detection, and denoising. The course aims to highlight the practicality and user-friendly nature of these deep learning techniques.
Also, a sponsored lecture on Apeer.com, Zeiss’s cloud platform, will focus on image processing, including machine and deep learning. At the end of the course, participants can actively practice segmentation and cell measurement using a virtual reality system from the same company.
This course is loosely related to Processing and Analysis of Microscopic Images in Biomedicine, we assume that the participant already has a basic understanding of image processing and is able to work with Fiji.
