MOTIVATION: Innovative microfluidic systems carry the promise to greatly facilitate spatio-temporal analysis of single cells under well-defined environmental conditions, allowing novel insights into population heterogeneity and opening new opportunities for fundamental and applied biotechnology. Microfluidics experiments, however, are accompanied by vast amounts of data, such as time series of microscopic images, for which manual evaluation is infeasible due to the sheer number of samples. While classical image processing technologies do not lead to satisfactory results in this domain, modern deep-learning technologies, such as convolutional networks can be sufficiently versatile for diverse tasks, including automatic cell counting as well as the extraction of critical parameters, such as growth rate. However, for successful training, current supervised deep learning requires label information, such as the number or positions of cells for each image in a series; obtaining these annotations is very costly in this setting. RESULTS: We propose a novel machine-learning architecture together with a specialized training procedure, which allows us to infuse a deep neural network with human-powered abstraction on the level of data, leading to a high-performing regression model that requires only a very small amount of labeled data. Specifically, we train a generative model simultaneously on natural and synthetic data, so that it learns a shared representation, from which a target variable, such as the cell count, can be reliably estimated. AVAILABILITY AND IMPLEMENTATION: The project is cross-platform, open-source and free (MIT licensed) software. We make the source code available at https://github.com/dstallmann/cell_cultivation_analysis; the dataset is available at https://pub.uni-bielefeld.de/record/2945513.

MOTIVATION: Microfluidic platforms for live-cell analysis are in dire need of automated image analysis pipelines. In this context, producing reliable tracks of single cells in colonies has proven to be notoriously difficult without manual assistance, especially when image sequences experience low frame rates. RESULTS: With Uncertainty-Aware Tracking (UAT), we propose a novel probabilistic tracking paradigm for simultaneous tracking and estimation of tracking-induced errors in biological quantities derived from live-cell experiments. To boost tracking accuracy, UAT relies on a Bayesian approach which exploits temporal information on growth patterns to guide the formation of lineage hypotheses. A biological study is presented, in which UAT demonstrates its ability to track cells, with comparable to better accuracy than state-of-the-art trackers, while simultaneously estimating tracking-induced errors. AVAILABILITY AND IMPLEMENTATION: Image sequences and Java executables for reproducing the results are available at https://doi.org/10.5281/zenodo.1299526. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Streptomycetes are filamentous microorganisms of high biotechnological relevance, especially for the production of antibiotics. In submerged cultures, the productivity of these microorganisms is closely linked to their growth morphology. Microfluidic lab-on-a-chip cultivation systems, coupled with automated time-lapse imaging, generate spatio-temporal insights into the mycelium development of streptomycetes, therewith extending the biotechnological toolset by spatio-temporal screening under well-controlled and reproducible conditions. However, the analysis of the complex mycelial structure formation is limited by the extent of manual interventions required during processing of the acquired high-volume image data. These interventions typically lead to high evaluation times and, therewith, limit the analytic throughput and exploitation of microfluidic-based screenings. RESULTS: We present the tool mycelyso (MYCElium anaLYsis SOftware), an image analysis system tailored to fully automated hyphae-level processing of image stacks generated by time-lapse microscopy. With mycelyso, the developing hyphal streptomycete network is automatically segmented and tracked over the cultivation period. Versatile key growth parameters such as mycelium network structure, its development over time, and tip growth rates are extracted. Results are presented in the web-based exploration tool mycelyso Inspector, allowing for user friendly quality control and downstream evaluation of the extracted information. In addition, 2D and 3D visualizations show temporal tracking for detailed inspection of morphological growth behaviors. For ease of getting started with mycelyso, bundled Windows packages as well as Docker images along with tutorial videos are available. CONCLUSION: mycelyso is a well-documented, platform-independent open source toolkit for the automated end-to-end analysis of Streptomyces image stacks. The batch-analysis mode facilitates the rapid and reproducible processing of large microfluidic screenings, and easy extraction of morphological parameters. The objective evaluation of image stacks is possible by reproducible evaluation workflows, useful to unravel correlations between morphological, molecular and process parameters at the hyphae- and mycelium-levels with statistical power.

MOTIVATION: Single cell time-lapse microscopy is a powerful method for investigating heterogeneous cell behavior. Advances in microfluidic lab-on-a-chip technologies and live-cell imaging render the parallel observation of the development of individual cells in hundreds of populations possible. While image analysis tools are available for cell detection and tracking, biologists are still confronted with the challenge of exploring and evaluating this data. RESULTS: We present the software tool Vizardous that assists scientists with explorative analysis and interpretation tasks of single cell data in an interactive, configurable and visual way. With Vizardous, lineage tree drawings can be augmented with various, time-resolved cellular characteristics. Associated statistical moments bridge the gap between single cell and the population-average level. AVAILABILITY AND IMPLEMENTATION: The software, including documentation and examples, is available as executable Java archive as well as in source form at https://github.com/modsim/vizardous. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.