Continuing our monthly interview series "Building a Free Internet of the Future", we spoke with the team behind the OpenFlexure open hardware microscope project, and in particular with Julian Stirling, Chief Executive of the Humanitarian Technology Trust. Now used in over 50 countries across every continent, the OpenFlexure project uses 3D printers and readily available components to produce open-source, laboratory-grade microscopes at a fraction of the cost of traditional models. Its use outside proprietary platforms has the potential to benefit many, with applications in fields such as pathology, education and training, telemedicine and much more.
"Building a Free Internet of the Future" is published by the Association for Progressive Communications (APC), highlighting the experiences and perspectives of individuals and communities supported by the NGI Zero (NGI0) grants. Funded by the European Commission, NGI0 supports free software, open data, open hardware and open standards projects. It provides financial and practical support in a myriad of forms, including mentoring, testing, security testing, accessibility, dissemination and more.
This interview has been edited for clarity and length.
Where did you hear about NGI0? And why did you decide to apply for a grant?
I first found out about NGI0 from the NLnet website, as I regularly check NLnet's grant calls. I first became aware of NLNet as a funder when my friend Kaspar applied for (and won) an NLnet grant for Open Know-How search. At that time I never considered applying with OpenFlexure, as I didn't feel that the microscope was within the remit of their funding.
More recently, I was one of the team members on their Open Hardware Manuals project. Working more closely with NLnet, I realised that their vision for the Next Generation Internet is actually very strongly aligned with the work OpenFlexure is doing by creating a microscope interface that complies with the W3C (World Wide Web Consortium) Web of Things standard, and with data-sharing of stitched microscopy images.
From “Microscopy for everyone” to “Reclaim the public nature of the internet”. How do you tie these two aims together?
When we think about the internet, we tend to think of the World Wide Web, but the internet is so much greater that that now. So much of modern life is internet-first, and so many of our interactions with individuals and companies are through online platforms. The OpenFlexure Community itself is a global community that has coalesced around an open design for a microscope, as a community that shares and builds together.
But as the internet has grown, we have seen some worrying walled gardens owned by huge corporations, hence the need for NGI0. One fast growing aspect of the internet, that is generally unseen to the general public, is telemedicine. This could be entirely remote appointments/diagnoses, but telemedicine also includes communication between doctors, allowing for remote consults and second opinions, enabling greater training opportunities for doctors in remote areas to join conferences discussing difficult cases.
In the case of diagnostic microscopy, for effective online communication you need specialised hardware – a robotic microscope that can automatically scan a sample and create huge giga-pixel composite. These proprietary machines are incredibly expensive, they take considerable infrastructure to maintain, but also they create images in proprietary formats that need further proprietary platforms for sharing. The OpenFlexure Microscope is a robotic microscope that can create similar huge composite images, but from an entirely open source hardware design, using entirely open source software, and saving the images in an open format.
Used in over 50 countries and on every continent, the project is used for pathology, learning and teaching, workshops, malaria diagnosis, microbiology and plants, water studies, etc. What are you going to do with the NGI0 grant?
The NGI0 grant is going to allow us to focus on the usability and stability of the software that drives the microscope. The microscope started as an academic project and most of the development has been done within academia. However, it is very hard to find time and funding within academia for user-centric design of a UI [user interface], for aspects of software maintenance, testing documentation, etc. – this is software development, not "research". A number of core technologies that the current version of the software is built on are out of date; we have an alpha version of a new server, but much of the work to date has gone into low-level code (as the camera stack we were using was deprecated!).
The NGI0 grant will not only allow us to make the next software release more reliable and with full features, but it will also help us to ensure that the interface is appropriate and accessible to our target audience, and to improve our processes so that future development is more sustainable.
Network access is required for “telemedicine” to work. This is a point of inequality in terms of access in different regions of the world and also a question of accessibility, which is often unfair. How are you working on this?
Yes, telemedicine requires the internet, but there are many forms of telemedicine requiring different levels of connectivity. We often think of live consultations where there is no doctor nearby and everything happens through an app or device. In many areas internet is available, but may be unreliable for real-time communication. However, if we consider a clinic with a microscope, if the microscope is digital and robotic, it can be left unattended to scan a sample and generate a stitched composite image; this image can be sent when internet is available, or placed on a device and taken to where internet is available.
It is important to stress that often, telemedicine for microscopy is not about off-loading or outsourcing diagnosis. Medical professionals regularly consult with their peers to get second opinions on rare conditions and difficult cases. In rural regions, doctors can be isolated from their peers, making second opinions difficult. The ability to get a remote diagnosis is transformative for patients, as it stops samples or even patients needing to travel long distances for a diagnosis, but it also means that there is no requirement for a stable enough internet connection to stream live microscopy, or to send all data collected from each scan.
OpenFlexure has been used in several countries for malaria diagnosis. In most places, access to high-quality microscopes is limited by the initial cost and maintenance expenses of the equipment. You are changing the game. What's more, it's an open hardware and free/libre software project. How is the project perceived and received outside the engineering bubble?
It's worth clarifying that, while we have shown that the microscope has the technical capabilities to image malaria parasites, we are still some way from the microscope being used for routine diagnosis. For routine diagnosis, we need to get to the point where local manufacturers can get the microscope certified for diagnostic use.
How it is perceived outside the engineering bubble is a great question. One fun thing about 3D printing is that you can easily make mechanical parts in almost any colour or even multicoloured with rainbow filament. You very quickly then realise that people take you far less seriously if you turn up with a bright pink microscope than a white one with blue highlights. First impressions are incredibly important, and something that it is easy to overlook when you are too close to the engineering side.
We have many enthusiastic collaborators in the medical space (some who do like multicoloured microscopes!), but I think the idea "clicks" with some people better than others. We will always look quite different from a traditional microscope or medical whole-slide imager, but I think we need to work on how we communicate and present the project to make it as familiar as possible to those who the technology is designed for.
What are the next major challenges for the OpenFlexure project?
The next major hurdle is getting the project to the point where manufacturers produce products certified for diagnostic use based on the microscope design. The more time we have spent looking into it, the more we see its value, even if it is a lot of work. It is one thing for a microscope to be able to see a malaria parasite, but it is quite another to ensure that every microscope that a manufacturer makes will be able to identify malaria parasites with the same fidelity routinely every time it is used.
Getting this correct takes a lot of thought about procurement, testing, calibration, self-calibration, self-testing, maintenance, wear and tear. It is essential work, some of which will need to be done on a case-by-case basis by manufacturers. But a lot of the thinking behind it is common to anyone manufacturing the microscope. I think it is quite exiting for an open source hardware project to go down the route of open quality control and open technical files, but it is also a lot of work!
The other huge challenge is working on the reliability and completeness of the software. Something that wasn't helped by Raspberry Pi deprecating their old camera stack; though we are happy that the new camera stack is far more open, so in the long run it is a good thing!
The floor is yours. Is there a message you would like to pass on to our readers?
Open hardware is not dead! There has been a lot of doom and gloom from some prominent open hardware projects. Gloom that hardware communities contribute back like software communities, doom that we are all going to be overshadowed by mass-produced Chinese clones.
Hardware is very different from software; the barrier to entry for contributions is much higher, and the distribution has huge logistical barriers and upfront costs. At the same time, there are lots of exciting things happening in the open hardware space, it is just a much younger community still exploring a much larger parameter space.
In the case of medical technology, there is a great need for collaboration and openness. We see some of this in academia, but as mentioned before, the work to transform a technology from a prototype stage to something that can be produced in quantity and trusted to perform for routine medical use is huge. This gap between prototype and market is normally called the “valley of death”; it is where most projects fail without huge investment that open source projects are generally not able to secure. It is for this reason that we have established a UK charity called the Humanitarian Technology Trust. The Trust is supporting the pathway to regulated medical production of OpenFlexure Microscopes; we plan to learn from this process and to apply it to more open source projects in the future.