An attempt to democratize light-sheet microscopy.
This is the repository for the open-source low-cost light-sheet microscope based on a monochromatic cellphone CMOS sensor and commercially available laser-scanning video projector.
Disucssion can also be found in Twitter
Everything on your own risk, no guarantees that it works or that it harms you! It's just a fun project without any professional use in mind!
- Low price (<500€)
- Portable
- Small footprint
- ready for the incubator
- Isotropic resolution of about 3µm
- Excitation Wavelength: 450, 530, 605nm
- Mechanical X/Y-Stage for coarse sample movevement
- Electrical X/Y-Stage for fine through-focus series of the sample
NEW: Now we support the system with a stand-alone Android App which takes care of the following:
- Image acquisition (monochrome Huawei, RAW, manual texp, ISO-settings)
- Hardware control of the X/Y-stage for through-focus series (MQTT-based)
- Control of the Light-sheet displayed on the Laser-projector
Everything can be found in the folder ANDROID. It has been tested with a HUAWEI P9. To display the pattern using a smartphone, one need to connect the laser-projector to the Phone using a HDMI-HML adapter (i.e. USB-c to HDMI), available for 8€ at Ebay.
The
The blue stripe is the secondary screen displayed in debug mode.
The
In this setup, we had to remove the lens to have bare accessoto the Sony IMX286 monochromatic chip. A minimum set of tools was necessary. A brief description with pictures is shown below
The Setup is relatively simple to assemble. A base-plate holds the Projector, the lenses (CCTV as objective and 25mm lens as illumination objective) as well as the mirror. The lid holds everything in place and the smartphone adapter keeps the cellphone in the correct distance from the CCTV lens.
One can use asimple syrringe for to prepare a sample e.g. in agarose. To reduce scattering, we provide a simple sample chaber which an be filled with water. The tip of the agarose can be put inside the water.
For the wiring of the two bipolar motors from the low-cost XY-stage, we used an Arduino Nano equipped with two H-bridges directly connected to the motor.
The following gives you a very brief introduction of the workin principle of the low-cost, but very precise Z-stage which is used in our setups. You can find a more detailed version of the description at aliexpress (chineses though). I have absolutely no clue where this piece is actually coming from, but it is working really great!
It is equipped with two bipolar stepper motors for X and Y direction. The wiring follows (from left to right):
1 -- A+
2 -- B+
3 -- A-
4 -- B-
It can be controlled with an ESP32 or Arduino for example. Code is very easy as it relies on the
The motor can be connected as follows:
If one uses an H-Bridge one simply need to make sure, that A+/A- and B+/B- are connected to MOTOR A and MOTOR B respectively.
- Impedance: ~45 ohm
- Stepsize: 18 degree
- Voltage: 5 V
- Peak Current: 100mA
- Movement X/Y: 2.4 mm
The code is from the Adafruit example with the wiring explained above:
#include <Stepper.h>
// change this to the number of steps on your motor
#define STEPS 200
// create an instance of the stepper class, specifying
// the number of steps of the motor and the pins it's
// attached to
Stepper stepper(STEPS, 4, 5, 6, 7);
void setup()
{
Serial.begin(9600);
Serial.println("Stepper test!");
// set the speed of the motor to 30 RPMs
stepper.setSpeed(60);
}
void loop()
{
Serial.println("Forward");
stepper.step(STEPS);
Serial.println("Backward");
stepper.step(-STEPS);
}
Make sure you're releasing the Motor after usage, otherwise it can get quiet hot!
In order to perform an experiment one need to do the following steps:
- Generate the light-sheet pattern using a computer by displaying a single-pixel line on a full-screen image (e.g. black screen, blue line (RGB=0,0,255))
- Eventually move the line periodically to find the best position of the light sheet (use the PPT in the presentation folder)
- Align the smartphone, so that the round-circle is in the center of the CMOS camera sensor (therefore take a light-source and shine light inside the smaple chamber, i.e. inline with the CCTV lens)
- Insert the syringe in the sample holder
- Move the X/Y move so that the sample is in focus by placing a light source perpendicular to the light-sheet or in-line with the detection objective
- Let the electric X/Y stage oscillate back and forth to acquire a through-focus series
- Acquire a video sequence of the sample
All files can be found in the folder STL.
- 1x Base Plate
- 1x Sample Chamber
- 1x Space for Smartphone
- 1x Lid for the Baseplate
- 1x Adapter for the XY-Stage: Arm which holds the syrringe
- 1x Adapter for the XY-Stage: Arm which holds the electronic XY-stage
(Links to the shops are not sponsored!)
- 1x cellphone (Huawei P9, remove the lens from the monochromatic camera) - 150€
- 1x XY-Stage (60x60xmm range) - 60€
- 1x motorized XY-Stage (aliexpress ) - 3€
- 1x ESP32 - 5€
- Set of M3 screws - 1€
- 3x Coverglass, 18x18mm - 1€
- 1x Laser-scanning projector, SONY MP-CL1A Link to Sony - 250€
- Filterset 1€
- Wires, USB-Cable, HDMI-Cable - 15€
- Comar Mirror d=10mm - 20€
- 1x CCTV Lens (f=4mm, f#=1.8) - 10€
- 1x 25mm lens, diameter: 15mm
The Huawei P9 does not allow the acquisition of monochromatic videos by default. Therefore we used the open-source app FreeDCam available from github or the playstore. In order to reduce the bouncing effect caused by intereference of the scan-speed of the laser (60 Hz), one need to select a frame-rate t_exp < 1/60 s respectively. ISO settings should be selected to get enough signal (usally 1000-2000) for high-quality samples.
Since Fiji does not accept .mp4
-files by default, one need to transcode the video into a sequence of images which can be imported through Fiji. This can conveniently be done by using ffmpeg
(freely available from Windows, MAC and Linux!). A standard command to extract the frames into a folder is the following:
ffmpeg -i MOV_2019_10_24_07_33_13.mp4 -filter:v "crop=1024:1024:512:512" -vsync 0 MOV_2019_10_24_07_33_13/cropped%06d.png
Here a lens-tissue "labelled" with fluorescent marker (Stabilo, Fluorescent Yellow) was swept through the focus. A theater filter blocks the light of the blue exciting laser.
A 3D projection of the result in ClearVolume looks like this:
Better images are coming soon!
Hardware follows the CERN license. Everything else is licensed under CC-BY-NC.
Everything is free to use - but we would like to ask you for any suggestions and comments. This makes sure, that we can improve this project.
Everything was developed in the @Heintzmann-Lab around Rainer Heintzmann at the IPHT Jena.
We need proper samples! If you know how to prepare easy to use cleared fluorescent samples ready for the SPIM, please contact us!! :)
In case you have any questions, please fiel an issue or write us an email. We try to improve the setup.
There is high potential to automate everything - if you have time and want to do that: We hire (but don't have any money - sorry ;-)