MORPHEUS: An automated tool for unbiased and reproducible cell morphometry.
Federico Alessandro RuffinattiTullio GenovaFederico Davide MussanoLuca MunaronPublished in: Journal of cellular physiology (2020)
Here we present a new Fiji/ImageJ2 plugin called Multiparametric Morphometric Analysis of EUcaryotic cellS (MORPHEUS), designed for the automated evaluation of cell morphometry from images acquired by fluorescence microscopy. MORPHEUS works with sampling distributions to learn-in an unsupervised manner and by a nonparametric approach-how to recognize the cells suitable for subsequent analysis. Afterward, the algorithm performs the evaluation of the most relevant cell-shape descriptors over the full set of detected cells. Optionally, also the extraction of nucleus features and a double-scale analysis of orientation can be performed. The whole algorithm is implemented as a one-click procedure, thus minimizing the user's intervention. By reducing biases and errors of human origin, MORPHEUS is intended to be a useful tool to enhance reproducibility in the bioimage analysis.
Keyphrases
- induced apoptosis
- machine learning
- deep learning
- cell cycle arrest
- single cell
- cell therapy
- randomized controlled trial
- endothelial cells
- endoplasmic reticulum stress
- high throughput
- cell death
- single molecule
- emergency department
- oxidative stress
- minimally invasive
- high resolution
- optical coherence tomography
- patient safety
- signaling pathway
- cell proliferation