Has anyone here had any success with a DIY slide scanner (no fluoro, just basic brightfield whole slide scanning)? I plan on overestimating my hobby time this year and am looking for a project to fail at. Plenty of protocols online, but love hearing the experiences of everyone here as well.
Not a lot of experience here, however I will offer a few thoughts.. For this purpose an inexpensive non-infinity objective from a company like Amscope should be âgood enoughâ for imaging (assuming high NA is not a goal). With an objective like this (or what is used in the OpenFlexure microscope) you can dispense with having a tube lens. A Raspberry Pi HQ camera will give you a pretty good RGB camera at a reasonable price. An LED array from Adafruit can be used if you are interested in trying different imaging approaches like DPC, fourier phytograpy or darkfield. The illumination is important, you will want a source that at least matches your objectives NA to get good images.
The hard part, and the cost differentiator is the mechanics. You need a fast and precise XY stage and focus stage to image an entire slide in a reasonable amount of time. I think the standard time (for 20x?) for a commercial scanner to image a 15x15mm area is on the order of a few minutes. This performance is difficult to match using hobby level hardware similar to what is found on a 3D printer or laser cutter.
Yes. I have had âsuccessâ with two builds. In one I motorised a Zeiss standard using simple A4988 drivers and a custom PCB:
The second one is something I am building now which is a âfrom scratchâ CNC stage for the PUMA microscope (total cost <ÂŁ1k, including microscope) see here:
The latter CNC stage is not quite finished but is at an advanced state of development. You can follow the progress on my YouTube channel âShortsâ and also âPostsâ:
as well as more detailed in-depth development videos and articles for my Patreon supporters:
Once the CNC stage module is ready for release I will be making public âHow Toâ videos and releasing the software but it is too early for that just now.
For a lower budget and easier build you could look into the OpenFlexure project. Although that is more limited in its abilities than my CNC stage, it may be that it provides what you need for basic small sample scanning and it is significantly cheaper and easier to make than my system (total cost about ÂŁ250 - ÂŁ300).
Then there is also the UC2 stage. Actually there are several âversionsâ of UC2 stages because they tend to build their stage around specific commercial XYZ motion mechanisms (like specific 3D printers or CNC machines from AliExpress) so, as one goes off the market they go ahead and design another one with whatever happens to currently be being sold.
Other options also exist - there are actually quite a few out there, each with its own strengths and weaknesses.