This repository provides a GUI application that facilitates the analysis of calcium imaging data - that includes EDA, cell identification, and extraction of summary statistics. The app is currently under development. The current version 0.15 runs on macOS (tested with High Sierra), but it should be Windows-compatible.
macOS High Sierra. Other OS might work as well.
If you want to build the app from source, make sure to install ($pip install <python_module>
) all dependencies that are listed below.
- keras and tensorflow backend
- skimage (scikit-image)
- numpy
- pandas
- matplotlib
- tifffile
- Tkinter
- tkMessageBox
- tkFileDialog
- pyobjc
Confocal microscopy images (a 'movie') of live cells are provide as an example ('data/'). The cells were grown on a coverslip for 48h and then incubate with a calcium dye for 30min. During the experiments, cells were stimulated with different concentrations of ATP to evoke calcium responses (i.e. increases in fluorescence intensity).
Currently, the following functionality is available:
You can load a file and use the 'Preview' function to see pixel value distribution and the impact of different filters.
A .lsm file exported from a Zeiss LSM series confocal microscope (e.g. LSM 710). The source code can be modified to integrate other file formats (e.g. .tif) as well.
'Preview' provides a pre-processing analysis of pixel value distribution and filters. Identification of cells is done via a connected components labeling algorithm. During the actual analysis, the identified cells are masked and tracked over time to derive a time course of relative fluorescence intensities.
Saving the analysis results to an output directory is currently not supported!
See figure titles and 'Analysis' section.