Illumination Correction

[roshankern]

In mitocheck_data we used BaSiCPy to determine the illumination function and temporal drift of the mitosis movies by providing BaSiCPy with the entire mitosis movie. However, when illumination correction on only a single frame is required, this method may not be worth the time it takes. Instead, it may be more viable to perform illumination correction on only a single frame. However, BaSiCPy needs at least 3 images to work and cannot take a single frame. Thus we have the following options for illumination correction during IDR stream with Mitocheck timelapse data:

1) Give BaSiCPy entire movie:
About 1.5 minutes, using BaSiCPy as intended.

2) Give BaSiCPy 3 frames of context:
Give BaSiCPy the frame of interest and 2 other frames (2 before or 2 after depending on frame of interest's position in movie).
About 1.5 seconds, no visible change in illumination correction.

See https://github.com/roshankern/IDR_stream/blob/add-IDR-stream/IC_test.ipynb for a visual comparison on methods 1 and 2.

3) Illumination correction for every batch/stream:
This method involves using multiple images from the IDR study (either a subset of the batch images or the entire screen) to derive an illumination correction function for the entire batch/stream.
Singh et al, 2014 say:

Retrospective methods that estimate an illumination correction function (ICF) by combining information across multiple images are more robust and thus more desirable... We have found that a straightforward approach to retrospective illumination correction works well in practice for high-throughput microscopy experiments. The approach is as follows. The ICF is calculated by averaging all images in an experimental batch (usually, all images for a particular channel from a particular multi-well plate), followed by smoothing using a median filter. Then, each image is corrected by dividing it by the ICF.

Peng et. al, 2017, the developers of BaSiCPy, say the following about this type of illumination correction:

These approaches take advantage of shared S(x) among an image sequence and are usually more reliable than single-image based corrections. Yet they require large numbers of images to reach a stable performance and a manual fine-tuning of internal parameters, and their robustness to common bioimage artefacts (such as dust and fluorescence dye particles) has not been tested. Moreover, none of the existing methods is able to model and correct temporal drift (e.g. caused by photobleaching) for time-lapse movies.

They go on to say about BaSiC:

In comparison to existing shading correction tools, BaSiC requires fewer input images, works for diverse imaging conditions and is robust against typical image artefacts. Moreover, it can correct temporal drift for time-lapse microscopy data, and hence improve single-cell quantification.

Note: Bioimage artefacts are not a major concern for Mitocheck mitosis movies as these have already been filtered out via Mitocheck quality control. However, these artefacts may be present in other IDR data.