Background correction - titration_plate_20200403_154849 - B01-0002

[tischi]

{"plateName":"titration_plate_20200403_154849","siteName":"B01-0002","pixelLocation":[2612.3864766701586,3207.0781571624098,0.0]}

Here, one cell seemed to have been dragged to the right, maybe during some washing step.

Interestingly, the intensity in the serum channel where the cell has been is dimmer than in the rest of the image. This indicates that the serum is binding to the coverslip.

Thus, if we subtract the background that we measure in the region outside the cells, we may in fact subtract a too high background value.

Screenshot

[tischi]

@metavibor @imagirom Please see above.

[metavibor]

@tischi @imagirom what is wrong with the idea to measure the "no cell area" for each well, calculate the median from all wells and subtract that value from each well? Than this issue of serum occasionally sticking to the well will not be relevant

[imagirom]

@metavibor The current approach is very close to what you propose: we take the median over all background pixels in the entire plate and subtract that.

The way I understand @tischi it that the lower background intensity in the area where the cell supposedly came from is closer to the background we should actually subtract (@tischi correct me if I am wrong, and maybe you can explain better).

[metavibor]

@imagirom @tischi think it is fine than... that lower background at the spot where the cell was should be pretty close to the median of the entire plate

[tischi]

The way I understand @tischi it that the lower background intensity in the area where the cell supposedly came from is closer to the background we should actually subtract (@tischi correct me if I am wrong, and maybe you can explain better).

That is correct!

In future plates, Vibor may also have one well without any serum added. This will be valuable to evaluate a minimal background value that we could use for subtraction.