![]() to describe always the same cell) is, with n being the length of the trace (i.e. a cell in one frame corresponds to the same cell in the previous frame) is constant in time, then the probability of the trace being correct (i.e. Assuming that the probability ρ of a correct cell assignment (i.e. Tracking quality decreases exponentially with the increase in the number of frames. Although such incorrect assignments are expected to be relatively rare at each time point, a simple analysis shows that the number of correct traces decreases rapidly with the duration of the experiment ( figure 1).įigure 1. two cells exchanged at some time point) can possibly hide interesting features or worse, create spurious information. However, the capability to extract single cell traces from microscopy images in a fully automated manner is a necessary prerequisite to obtain conclusions that are valid and biologically relevant in long-lasting experiments. Both aspects are of increasing importance to obtain a quantitative understanding of cellular processes at the single cell level. Moreover, microfluidics can be used to create time-varying environments. One can take advantage of microfluidic microchemostat that, unlike liquid bulk culture, enables long-term observations of cells growing as a monolayer. This is a decisive advantage to investigate a number of important biological problems, including chronological ageing, epigenetic heritability and dynamic features such as the cell cycle and circadian oscillations in non-synchronized cell populations. The former provides great statistical details on the diversity of the studied cell population, whereas the latter provides longitudinal information on single cells: individual cells can be tracked in time. Used in combination with fluorescence reporter techniques, flow cytometry and time-lapse microscopy are arguably the two most widely employed quantitative single-cell observation approaches. Observing cellular processes at the single cell level is often necessary to understand how cells respond to endogenous and environmental changes. As a community effort, we set up a website, the Yeast Image Toolkit, with the benchmark and the Evaluation Platform to gather this and additional information provided by others. We created a benchmark dataset with manually analysed images and compared CellStar with six other tools, showing its high performance, notably in long-term tracking. A graphical user interface enables manual corrections of S&T errors and their use for the automated correction of other, related errors and for parameter learning. The key features are the use of a new variant of parametrized active rays for segmentation, a neighbourhood-preserving criterion for tracking, and the use of an iterative approach that incrementally improves S&T quality. Here, we present CellStar, a tool chain designed to achieve good performance in long-term experiments. Surprisingly, even for yeast cells that have relatively regular shapes, no solution has been proposed that reaches the high quality required for long-term experiments for segmentation and tracking (S&T) based on brightfield images. With the continuous expansion of single cell biology, the observation of the behaviour of individual cells over extended durations and with high accuracy has become a problem of central importance.
0 Comments
Leave a Reply. |