To create a new ROI-based analysis for the currently loaded dataset, select the type of ROI you want to create from the "Add new" drop-down menu. Here, an analysis with automatically detected ROIs is created. This requires that F₀ estimation has been performed.

ROIs are discovered, based on local maxima in the fluorescence range projection (1). The fluorescence range image can be smoothed with a median and a Gaussian filter, via the "Sigma" and the "Median filter size" parameters. To deactivate one or both filters, you can set the respective value to 0 (zero).

By default, MSparkles uses a guided threshold. Therefore, an upper and lower threshold are computet, using Otsu's multi-threshold method. By adjusting the sensitivity slider, a final threshold is computed inbetween these two boundaries via linear interpolation. A sensitivity of 100% corresponds to the lower threshold boundary, while a sensitivity of 0% to the upper boundary.

You can test you parameters by clicking the "Test parameters" button at the bottom of the configuration window. This will generate a preview ROI-map (2), identical to the final result.