Open Source Cytometry Platform

3 minute read

This project is sponsored by Shapiro Lab. Dr. Shapiro concieved of the idea for an open-source static whole slide imaging cytometer and has published several articles in this regard:

  • Static cytometry in 1969 Science article
  • Static cytometry in 1994 “Cytometry” article
  • First pitch for static cytometry like outs in 2006 “Cytometry” article
  • 2014 pitch for this device in BioOpticsWorld article

Inexpensive technology such as the Raspberry Pi 3 and < $200 Basler scientific cameras have finally made Dr. Shapiro’s vision widely realizable.

Our initial prototype is a slide-based static fluorescence cytometer using whole slide 1:1 imaging and polychromatic sequenced LED stimulus. Our point-of-care diagnostic platform is intially targeted at the eradication of falciparum malaria with a $1 field test using a peripheral blood sample. Multiple characteristics of an entire slide are measured with a single camera and electronically-switched stimulus (no moving parts), exploiting high AT-pair percentage and hemozoin excretion of falciparum malaria.

Note that working documentation is on the project Wiki.

Open source cytometry conceptual block diagram. Imaging cytometer LED excitation module.
Static imaging cytometry prototype.

Executive Summary

Open-source/open-hardware development teams have successfully tackled important world problems and have stimulated large corporations to go open source as well–for example, NightScout for diabetes. We have developed a prototype point-of-care platform for cell analyses that can be built, used and maintained in resource-poor countries. This device is anticipated to be useful to count and classify blood cells and to detect and identify pathogens such as the parasites that cause malaria, which alone causes hundreds of thousands of deaths per year but which can easily be cured in days if diagnosed promptly. The bacteria that cause tuberculosis are another organism that this instrument may be used to detect and discriminate.

In affluent countries, blood cell analysis instruments use flow cytometry, a complex and expensive technology that needs highly trained operators and substantial infrastructure. Based on extensive experience developing such apparatus, we have partially prototyped a much smaller, simpler, and cheaper field-rugged instrument that can make the same proven measurements. The prototype open-source cytometry unit combines:

  • sub-$500 scientific-grade camera (typical smartphone chips lack quantitative characteristics)
  • inexpensive lenses and filters
  • pulsed high-power UV and blue LEDs
  • low-power green and red LEDs
  • embedded microcomputer for control, analysis, and the user interface.

UI choices include an onboard hotspot/webpage, reducing risks of built-in screens getting broken and enabling medical staff to get data directly on smartphones or tablets. The instrument, however, would be able to operate without an Internet connection. It should be able to be built for less than US$2,000 and run sustainably from solar cells and rechargeable batteries. The hardware/software platform developed would be immediately openly disclosed so that it could be built and used by clinicians and researchers from components widely available abroad. The range of applications extend beyond public health, e.g., to environmental and food science.

How it works

The device works by pulsing LEDs of known duration, spectrum and intensity onto a microscope slide with a dyed blood sample. The dye bonds to particular DNA pairs present in human cells and parasitic organisms also in the blood. Knowing the relative ratios of DNA pairs in benign vs. adverse organisms in blood, one can for example estimate likelihood of having a promptly fatal malaria case (~72 hours) or malaria that can be treated less emergently. Inexpensive sCMOS imaging technology and single-board computers with GPIO are key enabling technologies for this project.

Malaria types of interest for this project

  • Plasmodium falciparum - fatal within as little as 24 hours of symptom onset. Has 81% adenine-thymine pairs vs. human 60% A-T
  • P. vivax
  • P. ovale
  • P. malariae

Working code base

Intensity Ratio code demos finding cells and summing over the cell pixels in order to take intensity ratios. This is a post-processor, it does not control any hardware.

Raspberry Pi NoIR raw Bayer image acquisition code helped us see that cameras such as Pi NoIR can maintain ~ 1% intensity stability with our calibration. Stability is needed for quantitative cell image ratios re: stained slides.

LED illumination

The LEDs can’t be on too long as the dyes will bleach in a few seconds, making the measurements useless. Expected exposures are sub-second.