Explore chapters and articles related to this topic
Image segmentation
Published in Rodrigo Rojas Moraleda, Nektarios A. Valous, Wei Xiong, Niels Halama, Computational Topology for Biomedical Image and Data Analysis, 2019
Rodrigo Rojas Moraleda, Nektarios A. Valous, Wei Xiong, Niels Halama
Further validations are carried out using the software tool ilastik v.1.1.8 [41]. Ilastik is a general-purpose state-of-the-art image segmentation tool that provides an automated workflow based on the supervised training of a random forest classifier. For comparison purposes, training is performed manually on the same two images used to tune the proposed approach. The software is trained using color, edge, and textural features.
Principal component-based image segmentation: a new approach to outline in vitro cell colonies
Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2023
Delmon Arous, Stefan Schrunner, Ingunn Hanson, Nina Frederike Jeppesen Edin, Eirik Malinen
CellProfiler is another popular, free, open-source program that addresses a variety of biological features, including standard and complex morphological assays (e.g. cell count, size, cell/organelle shape, protein staining) (Carpenter et al. 2006). The program uses either standardised pipelines or individual modules that can be customised to specific tasks. Other macro-based colony detection algorithms implemented as ImageJ (Schindelin et al. 2012) plugins have also been proposed, such as IJM (Cai et al. 2011), Cell Colony Edge (Choudhry 2016) and CoCoNut (Siragusa et al. 2018). However, due to the sequential order of the modules, the performance of the cumulative operations may not be optimal on images from different experiments. Furthermore, a machine learning procedure has been combined with pipelines in CellProfiler to solve segmentation tasks – ilastik (Sommer et al. 2011). It uses a random forest classifier (Breiman 2001) in the training phase in order to assign each pixel’s neighbourhood into classes by interactive pixel labelling.