Hello everybody!

Many thanks for those who have registered interest in Cytkit, the open source spectral cytometer. It was great to meet so many enthusiastic people at recent events including CYTO and the Babraham Spectral Cytometry Symposium.

Cytkit really seems to have struck a chord among the cytometry community. Many people are interested for a range of reasons. The reason that I find most important is that cytometrists want power over what is in those boxes, just like they once did in the early days of cytometry. But in the intervening decades, commercial cytometers have gradually become fully-serviced black boxes. So an open cytometer uniquely gives the cytometrist new professional powers: to customise, design and tinker, to understand and investigate, and new tools to build new things.

With a couple of contributors, we are working to complete the Cytkit system, and gearing up for a pre-order campaign. The offer will go out when Cytkit is ready, and not before! Although the design so far (v0.3) is on Github, not many people can make use of it yet, because they would not only need to find all the parts, but also fill in the gaps: rustle up their own electronics, detectors and data acquisition system. So I want the design to be mature before we promise to deliver. That means at least another major iteration of the optics and mechanics. And, very importantly, a complete and well-documented electronics system and a fully functional software interface. Work is proceeding on all of these aspects.

Therefore, there is lots still to do! The pre-order is going to be delayed past September. Nevertheless, in the spirit of openness, I hope to take you along with us, by keeping you updated of our progress. There have already been a set of videos and technical posts on the website blog and posted on LinkedIn. I hope to keep up the cadence of a couple of updates per month until the pre-order campaign starts.

Feel free to get in touch if you want to contribute or have any thoughts to share!

Best wishes,

Samson Rogers

Open source optomechanics, following the Openflexure project. This is the second of a series of technical posts on the design of Cytkit, the open source cytometer.

As was pointed out by Richard Bowman, the founder of Openflexure, traditional optomechanics do not lend themselves well to open source 3D printed instruments. Each degree of freedom requires sliding joints (great in machined aluminium or brass, but poor in 3D printed plastic), plus toothed gears, rack and pinion. Moreover these parts need to be assembled, which adds to the cost and complexity. Instead, Openflexure pioneered flexure mechanics: by putting parallelograms into the structure, one could design several degrees of freedom in a single 3D-printed part [1].

Remarkably they reported less than 20 microns drift in their structure over one week, without temperature stabilisation [1]. Sounds ideal for many other applications of optomechanics…. actually I would venture to say that their ideas may be more powerful in cytometry than microscopy, because in a cytometer, one needs a few degrees of freedom and micron-scale precision of adjustment. However, one needs range of only a millimetre or so in each.

I set out to design optomechanics for a 3D printed cytometer based on their ideas. The first problem to solve was, what is the minimal number of degrees of freedom for this cytometer? Here we are talking about a single laser, forward/side scatter, and a single spectral array for a set of fluorescent markers. The answer is 5 – far fewer than in a commercial instrument. I can explain very briefly why. In a single 3D-printed body, we rely on the accuracy of the printer (~0.1 mm) for coarse placement of optical parts. Considering that the detector chips are each approximately 3 mm square, this accuracy is sufficient where the laser is not focused, However the laser focus itself is much smaller and has to be aligned accurately with the stream of cells, say within 10 microns. Since the stream is translationally symmetric in one direction, that means 2 axes are required for laser positioning, i.e. the laser focus lens is on an XY platform. Then the fluorescent light coming off the cells has to be collected and directed to an array of the detectors. One axis is required to focus the condenser lens, and two to point the collimated light in the right direction towards the grating and detector array. Considering that it is more compact to have 2 flexures in the same place rather than 3, the condenser lens is on an XY platform, while the laser is on a tilting flexure.

That totals 5 degrees of freedom: two XY stages and one tilt. The current design can be downloaded on the Cytkit repo on Github – follow links from the website in comments below.

Please sign up on the Cytkit website if you are interested in buying a kit once released, link below. We will be launching a pre-order campaign soon.

Get in touch if you want to talk about Cytkit, life science tools, or need my help for anything!

How to make an open-source sheath? This is the first of a series of technical posts on the design of Cytkit, the open source cytometer.

Almost all the fluidics parts needed for a flow cytometer are available as low cost generic components; most of the mechanical parts are also easy to 3D print. But not so the sheath chamber and cuvette. The development version of Cytkit uses a glass capillary instead of a quartz cuvette, since the former is widely available*. And that leaves the sheath part.

Current leading commercial cytometers mostly use a sheath chamber integrated with the cuvette. This is made of several machined and polished parts of fused silica, diffusion bonded, and available from a couple of specialist companies. These are far from cheap or widely available!

There is very little theory of sheath flow in flow cytometry, at least not in the public domain. What little exists is in the proprietary designs of the flow cytometry companies. Fundamentally the sheath has to make laminar flow: concentric sheath fluid and sample inputs, with the sample input well centred. And it should have nothing to disturb the sheath flow or bias it to one side, e.g. it should not trap bubbles. It should have smooth walls and avoid background fluorescence. A couple of paragraphs about sheath flow can be found in Howard Shapiro’s classic Practical Flow Cytometry. There is not much more theory around than that!

I designed a sheath chamber for 3D-printing. taking advantage of resin printing**. (Earlier versions used FDM, although the resolution wasn’t high enough and it was always tricky to get them to seal.) The latest version is a beauty, even if I do say so myself! The design is *very* minimalist: it has barbs for easy connection of silicone tubing, avoiding junctions (which trap particles) and avoiding microfluidic connectors (unnecessary cost).
The current design can be downloaded on the Cytkit repository on Github – follow links from website in comments below. See the video… it works! Aptly, I have rigged up an Openflexure microscope, custom mounted on its side, to verify the sheath flow by imaging of beads.

Get in touch if you want to talk about Cytkit, life science tools, or need my help for anything!

* For the potential performance advantages, I’m still considering adding a low-end quartz cuvette, even if it does have to be sourced from specialist companies.
** For the kits, we are considering using a specialist 3D resin-printing microfluidics service. Thanks to Paul Marshall of Rapid Fluidics for advice.

Here’s a teaser video about Cytkit now in development, showing off the minimal optical design and the fully 3D-printed mechanics

Is it possible to make a sub-$5k open-source cytometer that is functionally equivalent to a $100k commercial instrument?

Furthermore, is it possible to build a business selling such an open-source cytometer (as components, kits and fully assembled instruments)? No commercial organisation, no service contracts. The design needs to be minimalist and open-source to be accessible to the users, to install and configure for themselves, maintain and repair. Moreover, the business would rely on the community for the sharing of knowledge and design modifications for user applications.

It would also rely on an untapped resource – the cytometrists. In the past, cytometrists were the experts on building and modifying their hardware. However, this diminished greatly as automated fully-serviced commercial cytometers came to dominate the field. Cytkit would be a platform for the cytometrists to take back the field, to build and customise instruments for their users, enabled by open-source (including the optical model), minimalism of the basic design, community knowledge, and readily available components. So the cytometrists, instead of just charging their users by the hour for access to their instruments, could also provide a service to set-up and maintain open-source, custom instruments for their users.

I just came back from CYTO in Denver, Colorado, where I presented this concept, at David Novo’s Open Cytometry Hardware meeting and at my poster. I spent two days carrying the Cytkit prototype in a little flight case, showing it off to everyone who was curious to see it. People were excited, and now I’m now excited by their responses. Thanks to those who have already signed up on the website or told me privately that they will buy one.

So I hope the answers to the questions above are “yes”. I’m embarking on this business as an interesting experiment. I hope to start a pre-order campaign circa September, once the design is mature. Until then, look out for further posts about the development.

Please subscribe to the Cytkit LinkedIn page, or sign up on the website for email updates (link below).