Embedded Files
A full OpenSpin setup allows acquisition of both LightSheet and optical tomography (aka OPT) datasets. However, if you are only interested in optical tomography an OPenT is much simpler to assemble and operate. Light-sheet performs optimally with magnifications ~10-20x and with samples up to 1mm. OPenT is better suited for samples which need low magnification lenses (<4x) and samples a few mm or even cm. A simple OPenT scanner can be built in less than 1day, with the possibility of operating with transmitted light (a diffused while light source standing behind the sample) + 1 channel fluorescence emission mode (blue [LED] incident light & green fluorescence detected by the camera). This very basic OPenT did not include shutter or filter wheels, but is sufficient for acquired high-quality datasets (see for example the Haeckliens site)
NEW!
For samples labelled with two (or more) fluorescent dyes, or light-sensitive fluorophores, we redesigned the OPenT to include the necessary sub-systems for illumination with multiple LEDs (Eg a CooLED unit) while keeping the system simple and small footprint; we also prepared a new plugin (OPenT) with only the functions for OPT, and a macro to pre-process the datasets for filtered back-projection reconstruction (eg with the freely available nrecon software). This makes it much simper for those interested only in a high-end OPenT system; we have also optimized the hardware control and acquisition times are ~5x faster than the original.
The source code for this can be downloaded from here: https://github.com/hmmpereira/OpenSpin
For mode details on parts, assembly, new plugin or macro please contact gaby@igc.gulbenkian.pt
(the following instructions and diagrams relate to the version we published before -> Gualda et al 2013 and Felix et al 2016 [please cite us!])
Diagrams
Details on the parts list can be found here!
All parts should be assembled onto the breadboard according to the following diagrams:
(detail of sample block [with specimen embedded] submerged in the cuvette filled with BABB and illuminated with blue incident light)
DIY OPenT sample chamber (cuvette)!
Start with a common glass slide (75mm x 25mm x 1mm) and using a diamond knife cut it into 3x25mm squares. Buff the sharp edges to prevent accidental injuries during assembly and use!
Using another glass slide, use it as the base of the chamber and then glue the three walls in the center as in the figure sequence below (cyanocrylate glues ARE NOT stable enough once exposed to BABB so avoid!). We have seen inconsistent resistance to BABB from brand to brand of cyanocrylates, epoxy and silicone glue. Try! Locktite cyanocrylate seems to work well but once opened it loses eficacy rather quickly, so buy small instead of large bottles.
We had more luck lately with RTV silione glues [especially the "neutral" ones, not those that smell like vinegar] for cuvete assembly. You may find it also at auto-stores, ask for a silicone glue for the engine cooling system.
Finally add the front "wall" using a cover glass (22x22x0,17mm). This thinner wall of glass will face the camera+lens. The larger base (75mm) facilitates handling of the cuvette and avoid fingerprinting the lateral walls. Be extremely careful with the "front" window, as the coverglass is fragile and cracking it could mean leaking BABB and possibly ruining the breadboard or other components. Larger cuvettes can be similarly assembled starting from large microscope slides/cover glass available from microscope suppliers (often expensive, though!). During imaging, we maintain the cuvettes inside a large glass Petri-dish covered at the bottom with absorbing tissue, to contain any accidental spillage of immersion medium. The commercial "cells" (eg Helma) are expensive but highly durable and resist all solvents. They have sizes that are not available in the catalog...ask!
Detail of sample chamber front, facing towards camera lens:
click on the images to enlarge
Detail of incident light assembly (LED + optics) facing sample chamber;
Positioning the stage and camera:
The XYZ stage holder and kinematic mount attached to the motor facilitate positioning, centering and leveling of the motor axis in relation to the camera center-view. This is ABSOLUTELY ESSENTIAL for proper image reconstruction, as even slight deviations of just a few decimal degrees will case major aberrations in the final reconstruction. Start with a 2-4cm flat cap screw connected to the center of the magnet, dip it into sample chamber with camera turned on and adjust XYZ and camera to get the screw in the field-of-view (FOV). Leveling of the axis can be facilitated also using a "3 axis bubble level", such as those sold in photography stores (eg http://www.amazon.com/Polaroid-Triple-Bubble-Digital-Cameras/dp/B005FRI50K/ref=pd_sim_p_7), glued to a 3cm diameter metal washer and connected to the magnet (be careful, BABB will ruin the plastic!).
The camera should also be leveled, and the front wall of the sample chamber (cuvette) should be perfectly parallel to the objective front. Even after these adjustments, minor corrections may be necessary a posteriori as explained in the Supplementary Materials. When the specimen is longer than wider (eg an embryo) the camera should be rotated 90º to fill the camera chip (in the screen, the embryo will be rotating horizontally; before back-projection reconstruction, images need to be rotated again so that the axis of rotation is vertical!).
Adjusting the illumination:
Transmitted light source: center the light-bulb with the middle of the back wall of the cuvette and ensure that the FOV is evenly illuminated in the image. To hold the diffuser glass upright, glue it to a metal plate. If FOV is uneven, acquire an image without sample and use it later to subtract the background.
Incident light source (fluorescence): the source should be placed as close to the sample as possible ensuring that it is fully illuminated. Try to illuminate from an angle as close to the detection axis as possible (without blocking the field view) and avoid the cuvette corners. Larger samples will require moving the LED source away as possible to ensure full illumination and will require longer exposure times.
Note: There are in-axis options commercially available for incident light; it requires collimation of the LED light and additional optics which are quite expensive, and in our experience the light losses are considerable in some cases. All images we have produced or published so far were done with off-axis illumination as described above, for samples from a few mm to cm thick.
Positioning the sample:
After embedding the sample in agarose, dehydrating and clearing (see Supplementary Materials) glue the agarose block to a 3cm diameter metal washer using high-quality cyanocrylate glue, press gently and wait a few seconds. If block is smaller than hole in the washer, first glue a coverglass to the metal washer to serve as base. (Attention, make sure you use metal washers of a larger diameter than the magnet, otherwise centripetal magnetic forces will make it difficult to center the sample to the axis of rotation!).
Lift the motor as far as possible using the Y translation of the XYZ stage holder and, using forceps, invert the block and bring the metal washer part to contact with the magnet, immediately submerging the block to prevent the formation of bubbles. Using transmitted light, move the XYZ stage controls to bring the sample close to the front wall of the sample chamber (cuvette). Bring camera to focus and lock the rail carrier. Using forceps try to slide the blocks so specimen is as close to the axis of rotation as possible, first by looking from the front, then sideways. Minor focus adjustments can then be done either with the Infinitube standard focusing knob, or by moving the Z translation of the XYZ sample stage. While the motor rotates (use the function "lines" in the OPEN SPIN plugin window, and rotate the sample 360º), gently push the metal washer using forceps so that the center of the specimen coincides as much as possible with the yellow cross-hairs in the middle of the FOV. This is often a tedious trial-and-error task, but necessary! If the specimen is off-axis it will orbit and fall off the field of view occasionally. Also, because low magnification lenses are not perfectly flat-field corrected, image quality degrades away from the center of the field of view. When the specimen is finally aligned with the axis of rotation proceed to acquiring a full projection dataset and then reconstruct, as per instructions in Supplementary Materials of our original 2013 article.
OPenT Image Example
More examples here: www.gabygmartins.info/haeckaliens
(Mouse embryo E14.5, green auto-fluorescence; Left panel is a series of sagittal sections, right is a 3D volume rendering)
Page updated
Google Sites
Report abuse