miniscope V4 lens configurations

[Miguel Carvalho]

Hi, 

I am currently using V4 miniscopes to register neuronal activity through a prism+GRIN lens (inscopix) implant. Upon signal extraction I am observing a majority of units with unexpectedly sparse activity, even taking into account the physiological differences between spikes and calcium transients. In my preparations I have a dense cell labelling, with likely a high degree of overlap between cells. 

I am thinking that this, combined with the large FOV of the V4 lens configuration I am currently using (lens configuration 1 - 2x 3mm diameter, 6mm FL achromat) could be affecting the quality of signal extraction, by for example assigning to different cells "spikes" that are in fact from one cell. Spatial downsampling steps routinely implemented by 1P extraction algorithms could play a role here, as well.

I wonder if my preparation could benefit from a different combination of lenses that could provide a better spatial resolution, even at the expense of a larger FOV or higher working distance.

Thanks,

Miguel 

[Daniel Aharoni]

Hi Miguel,

While there is a chance decreasing the FOV and increasing the spatial resolution would help, the standard V4 Miniscope optics/resolution should be able to handle most (if not all) densely labelled brain regions. You it be possible for you to share a short example recording so we can get an idea of how dense and overlapping the cells in your data are?

What analysis pipeline are you using? There is a chance that some of the spatial parameters in the pipeline need to be adjusted to match the spatial resolution of the V4.

Finally, it is possible to swap out one or both of the achromatic lenses in the objective module, or to swamp out the achromatic tube lens in the Miniscope body, to increase the spatial resolution of the system. I think you can go to one or two 4.5mm FL lenses for the objective and see where that gets you. This will also have the effect of decreasing the FOV and working distance of the system. 

[Miguel Carvalho]

Hi Daniel,

Thanks for the quick reply. I have sent you an email with a link for one of my recordings.

Currently I have been using this analysis pipeline (https://bahanonu.github.io/calciumImagingAnalysis/). I apply a spatial downsampling factor of 4, though it is possible to reduce it. The limiting factor then becomes computer memory, as lower downsampling quickly uses up all the memory available. I am in touch with the developer of the pipeline to get some hints on how to optimize signal extraction, but I'll be glad to get any tips you can give me based on your personal experience or your assessment of the sample video I sent you. 

I am glad to know that there's a posibility to use a combination of lenses with a lower FOV. Are these the ones you mentioned? 

https://www.edmundoptics.com/p/3mm-dia-x-45mm-fl-mgfsub2sub-coated-achromatic-doublet-lens/7872/

If I am to assemble a new objective module using one 6mm FL lens and one 4.5mm FL lens should they be placed inside the objective module in a specific arrangement? 

Best,

Miguel

[Daniel Aharoni]

Hi Miguel,

Thanks for sending me an example video. I think the native spatial resolution of the V4 should be totally fine for resolving cells and their individual dynamics. It is somewhat likely that the 4x spatial downsampling you are using in your analysis pipeline needs to be removed. The V4 Miniscope is able to achieve its large(ish) FOV by having a lower internal magnification compared to the previous V3 version and pretty much any other open-source or commerical miniature microscope on the market. This effectively acts as spatial downsampling within the optics, making better use of the pixels available, and giving you a larger FOV. Some additional downsampling during analysis is fine but this will be dependent on a case to case basis. 

If memory is an issue, consider:

• Cropping the original Miniscope video to decrease the number of pixels uses in the analysis
• Trying a different pipeline. I think CaImAn and Minian for Python have reasonable memory demands
• Temporally downsample more

[Miguel Carvalho]

Hi Daniel,

Thanks for the explanation. I will test your suggestions and see if that gives me better results. 

Just a final question. In case I decide try out lan objective module with a lower FOV, could you confirm these are the correct ones? https://www.edmundoptics.com/p/3mm-dia-x-45mm-fl-mgfsub2sub-coated-achromatic-doublet-lens/7872/ 

Should the 6mm FL lens be placed inside the objective module first and then the 4.5mm FL lens?

Thanks for your help!

Miguel

[Daniel Aharoni]

Hi Miguel,

Yes the lenses you linked to look correct. Generally speaking, you should put the lens with the longer FL closer to the brain. A word of warning though, I do not think we have ever assembled a V4 with a 4.5mm and 6mm FL lens objective (or a 4.5mm and 4.5mm FL lens objective). I am confident it should still give you decent imaging but just wanted to make it clear that this hasn't been directly tested by us yet.

[Miguel Carvalho]

Hi Daniel,

Thanks for the additional input. If I test it out I will let you know how it goes.

All the best,

Miguel