Photographing whole slide-mounted specimens
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Photographing whole slide-mounted specimens
I dread to think how many hours I've spent at the microscope photographing and stitching tiles to make images of entire slide-mounted specimens. The lowest power objective on my Zeiss scope is 2.5x which covers a FoV about 4mm wide. Even modest-sized specimens demand lots of images to make a stitched panorama of the subject. Multiples more if it's focus stacked.
Until now, it honestly never occurred to me to treat that as a macro task. I always went straight to the scope and trudged through the process. Well, not any more I don't...
Attached is a longitudinal section of a rat heart, about 18mm long, imaged with an MP-E 65mm at just under 2x. Only one shot. The slide was backlit by flash-illuminated white paper with the slide held a few inches in front of that. I only used one flash and got a bit of a gradient on the background, but you get the idea. So trivially easy! I can't believe I never thought of it before. Duhh! Go on, laugh at me if you want...
Close-up included for decoration - taken with a 40/0.75 Plan Neofluar. The FoV is ~250 microns wide.
Until now, it honestly never occurred to me to treat that as a macro task. I always went straight to the scope and trudged through the process. Well, not any more I don't...
Attached is a longitudinal section of a rat heart, about 18mm long, imaged with an MP-E 65mm at just under 2x. Only one shot. The slide was backlit by flash-illuminated white paper with the slide held a few inches in front of that. I only used one flash and got a bit of a gradient on the background, but you get the idea. So trivially easy! I can't believe I never thought of it before. Duhh! Go on, laugh at me if you want...
Close-up included for decoration - taken with a 40/0.75 Plan Neofluar. The FoV is ~250 microns wide.
Good idea.
I have done it in the past with petrological thin sections and polarized light.
Another method I've done is to put a biological slide into a Nikon Super Coolscan 4000ED film scanner with a home made adapter (Nikon made this accessory)
Of course in both cases you can't zoom too much because resolution is limited.
I have done it in the past with petrological thin sections and polarized light.
Another method I've done is to put a biological slide into a Nikon Super Coolscan 4000ED film scanner with a home made adapter (Nikon made this accessory)
Of course in both cases you can't zoom too much because resolution is limited.
Pau
Indeed, but unless you're going to print it the size of a barn, it's resolution enough. The heart pic above was 42 megapixels before resizing for upload (minus small crops top and bottom). Resolution at pixel level is good, so a 40-inch print is feasible even with this single shot.Pau wrote: Of course in both cases you can't zoom too much because resolution is limited.
The required resolution of the output is something I've long neglected - I just jumped right in trying to get the very best resolution at the input. Wasted effort! Working with higher resolution cameras in recent years, and the subsequent slugging of my PC trying to deal with the huge files (especially with deep stacks) kind of brought the issue to my attention. Now I always consider the output resolution first, then adjust the input accordingly. Another thing I probably should have been doing all along - but better late than never. And a lesson always sticks better if you learn it for yourself - the hard way
It’a very good point. Lots of stitching and stacking to post a final 300kb, 1024 x 800 image is rather a waste of effort. The full image of the heart is very nice but it would also be nice to be able to zoom in and see the amazing detail that the 40x objective captures. As download speeds get faster, storage gets cheaper and automated stack-and stitch-gets easier uploading multi gigabyte sticthed images may become quite common - I hope so.
post subject
What about applying the Zoomify idea that micro_pix is using in photography through microscope.
good idea of yours, and impressed with the results.
good idea of yours, and impressed with the results.
used to do astronomy.
and photography.
Zeiss Universal Phase contrast.
Zeiss PMII
B&L stereo zoom.
and photography.
Zeiss Universal Phase contrast.
Zeiss PMII
B&L stereo zoom.
I think it would take well over 2000, maybe 3000, photos to get a whole image of the heart at 40X - without any stacking.
The final jpg I produced after stitching 14 images of the plant stem was 22.7mb 11150 x 8831 pixels and the Zoomify files are 20mb
I suspect a 40x stitch of the rat's heart would be 4-5GB.
It sounds doable Beatsy - if you haven't got anything better to do over the next few days.
The final jpg I produced after stitching 14 images of the plant stem was 22.7mb 11150 x 8831 pixels and the Zoomify files are 20mb
I suspect a 40x stitch of the rat's heart would be 4-5GB.
It sounds doable Beatsy - if you haven't got anything better to do over the next few days.
Last edited by micro_pix on Wed Mar 07, 2018 10:30 am, edited 1 time in total.
Re: Photographing whole slide-mounted specimens
Back when I was at Oxford I spent 9 months imaging mouse brain sections at 20x on a very high end Zeiss microscope. Those didn't look half as good as your image.
I don't own the images as they are published but you can see them here:
https://media.springernature.com/m685/n ... 0-SF10.jpg
https://www.nature.com/articles/nn.4250
I don't own the images as they are published but you can see them here:
https://media.springernature.com/m685/n ... 0-SF10.jpg
https://www.nature.com/articles/nn.4250
"recoding a spatial memory engram can alleviate associated maladaptive behavior."perdu34 wrote: https://www.nature.com/articles/nn.4250
When do you start the human trials?
Please don't apply it to the members of this group - we'd fail the testmicro_pix wrote:"recoding a spatial memory engram can alleviate associated maladaptive behavior."perdu34 wrote: https://www.nature.com/articles/nn.4250
When do you start the human trials?
I'm sure we can find a few good souls who want to help push the boundaries of science here.Beatsy wrote:Please don't apply it to the members of this group - we'd fail the testmicro_pix wrote:"recoding a spatial memory engram can alleviate associated maladaptive behavior."perdu34 wrote: https://www.nature.com/articles/nn.4250
When do you start the human trials?
Looks incredible, this is one of the things I am trying to get set up on my microscope to basically automate the stitching process by setting the top left and bottom right positions and getting it to run through and capture images bit by bit. Doing it by hand would be a nightmare.
Lower resolution, but depending on the video mode of your camera you could also try and record a view of you panning through the slide in a systematic way (e.g. zig zag) and then you could try and stitch this using Microsoft image composite editor.
Lower resolution, but depending on the video mode of your camera you could also try and record a view of you panning through the slide in a systematic way (e.g. zig zag) and then you could try and stitch this using Microsoft image composite editor.
Great tip. I too have done a few stitches of large specimens with a 2x nikon objective. Your point is a good one as on my Diavert setup, the final FF image is another 3.5x larger than the same magnification on my macro setup.
I've done a mouse heart. I also was able to borrow a human fetus longitudinal cross section which is very interesting. Doing these large specimens on my macro setup would have been a big time saver.
On a side note, it would be great to get into making thin sections and learning how to properly stain and mount them. I got a tour of the process at a local lab, and they have an enormous amount of machinery and techniques available to do this all properly. Just seems too daunting and expensive for the home hobbyist.
Speaking about properly mounted and stained specimens, where do you all purchase well done slides? I've bought a few slides from Carolina biological, but they didn't seem prepared nearly as well as the slides done at this local lab.
Steve
I've done a mouse heart. I also was able to borrow a human fetus longitudinal cross section which is very interesting. Doing these large specimens on my macro setup would have been a big time saver.
On a side note, it would be great to get into making thin sections and learning how to properly stain and mount them. I got a tour of the process at a local lab, and they have an enormous amount of machinery and techniques available to do this all properly. Just seems too daunting and expensive for the home hobbyist.
Speaking about properly mounted and stained specimens, where do you all purchase well done slides? I've bought a few slides from Carolina biological, but they didn't seem prepared nearly as well as the slides done at this local lab.
Steve