Scintillate: An open-source graphical viewer for time-series calcium imaging evaluation and pre-processing

- Matlab-based graphical user interface for evaluating time-series image kinetics.
- Allows whole matrix, frame-by-frame pixel change comparisons.
- User may perform ICA to separate independent signals from the time series.
- User may perform pre-stimulus background subtraction and additional image processing.
- Open-source. Deployable on Windows and macOS as a stand-alone application.

BackgroundCalcium imaging is based on the detection of minute signal changes in an image time-series encompassing pre- and post-stimuli. Depending on the function of the elicited response, change may be pronounced, as in the case of a genetically encoded calcium-reporter protein, or subtle, as is the case in a bath-applied dye system. Large datasets are thus often acquired and appraised only during post-processing where specific Regions of Interest (ROIs) are examined.New methodThe scintillate software provides a platform allowing for near instantaneous viewing of time-sequenced tiffs within a discrete GUI environment. Whole sequences may be evaluated. In its simplest form scintillate provides change in florescence (ΔF) across the entire tiff image matrix. Evaluating image intensity level differences across the whole image allows the user to rapidly establish the value of the preparation, without a priori ROI-selection. Additionally, an implementation of Independent Component Analysis (ICA) provides additional rapid insights into areas of signal change.ResultsWe imaged transgenic flies expressing Calcium-sensitive reporter proteins within projection neurons and moth mushroom bodies stained with a Ca2+ sensitive bath-applied dye. Instantaneous pre-stimulation background subtraction allowed us to appraise strong genetically encoded neuronal Ca2+ responses in flies and weaker, less apparent, responses within moth mushroom bodies.Comparison with existing methodsAt the time of acquisition, whole matrix ΔF analysis alongside ICA is ordinarily not performed. We found it invaluable, minimising time spent with unresponsive samples, and assisting in optimisation of subsequent acquisitions.ConclusionsWe provide a multi-platform open-source system to evaluate time-series images.