V4 imaging problem

[Dorien Maas]

Dear miniscope community,

We have just managed to image the first few blinking cells!
However, we get a strange light effect from the miniscope itself. I attach a picture.
The middle is super bright, which messes with imaging and image analysis. This also happens when we just image our surroundings and so it must be something with the miniscope itself.

Does anyone know how I can get rid of this? For example, did I make a mistake in miniscope assembly?

Thanks a lot for your input! 
Dorien

[William Mau]

The bright background in the middle is normal. It's usually corrected post hoc. For example see: https://minian.readthedocs.io/en/stable/pipeline/notebook_2.html#glow-removal-and-visualization

--
You received this message because you are subscribed to the Google Groups "Miniscope" group.
To unsubscribe from this group and stop receiving emails from it, send an email to miniscope+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/miniscope/1d7f9b8e-4a60-4bd4-9538-c350ce1269fan%40googlegroups.com.

[Dorien Maas]

Thank you for your answer!

I've tried correcting it with Minian analysis. However, the bright spot is much brighter then our fluorescent cells and so it makes the video's impossible to analyse, the only seeds detected are parts of the bright spot. Is there a way to reduce it?

Dorien

[Zeeshan]

Maybe playing with the gain level or LED strength could reduce this (while still picking up cells). You could try a few short sample recordings with different gain and LED levels and see if any of them improve cell extraction. 

To view this discussion on the web visit https://groups.google.com/d/msgid/miniscope/76428355-2c72-43f7-a8b3-da708b4220a4n%40googlegroups.com.

[Daniel Aharoni]

Hi,

What you are seeing there is a very small bit of excitation light leaking through the Miniscope and making it all the way to the image sensor. To some extent, this is somewhat unavoidable due to the excitation light coming from the excitation LED not being fully collimated. Optical filters are designed ideally for collimated light passing through exactly perpendicular to the filter's surface. Since this isn't really possible to achieve in the small form-factor of miniature microscopes, these filters perform somewhat sub-optimally. 

This is generally a non-issue but at very high LED intensities and image sensor Gains, the excitation leakage can become visible. There are a couple ways to deal with this:

• Much of this leaked excitation light makes its way to the image sensor by reflecting off the side surfaces of the dichroic filter. If you take this filter out and paint all its sides with black enamel paint or black optical paint it should drastically reduce the light leakage you see.
• A somewhat easier option is to take your fully assembled Miniscope and place it directly on top of a completely non-fluorescing surface (most materials will fluoresce a bit under a lot of blue light). A commonly available surface that doesn't seem to fluoresce is black electrical tape. Next, turn up the LED and Gain to around the settings you usually use and then hit 'space bar' in the Miniscope software to save an image from the Miniscope video stream (or just record a short video and then extract a frame from it). This image/frame shows just the distribution of light leakage through your system. Finally, whenever you do processing or analysis on your data, just subtract out this image/frame from each from of your dataset. This should effectively remove the light leakage from your data and allow you to continue on through processing and analysis. Just make sure you generally are not saturating the image sensor in this bright region during recording as that will end up resulting in a loss of data.

[Dorien Maas]

Great!! Thanks a lot for your answer, I'm going to try both options.

Dorien

[Florian Steenbergen]

Hi Dorien,

how did you end up handling this? Did you try to fix it on the hardware side or did you continued with the image substraction approach?

Would be super happy to hear from your experience, as we seem to have the same behavior in a miniscope of ours.

Thanks a lot and best,

Florian