Shipped medical device · In labs worldwide · Led to DP600 follow-on commission
🔗 Official product page. 🔗 My case study
Roche’s VP asked for a 1-click scanner interface. We designed one that needed 0 clicks. The VENTANA® DP200 is now in labs worldwide and there’s an LED panel on the hardware enclosure that wasn’t in the original brief, because we spent enough time in the labs to know it needed to be there.
The product
The VENTANA® DP200 is a digital pathology scanner from Roche Diagnostics, converting tissue slides into high-resolution digital images for clinical and research use. It is now in labs worldwide.
The brief
This was 2016. I was Lead Designer at Impossible, working alongside a colleague, consulting for Roche Diagnostics on a new scanning platform for small-batch lab jobs.
The state of the art for these devices was grim: step-based workflows, Windows XP-era interfaces, complicated flows requiring formal training before a lab technician could be productive. The competitor landscape wasn’t meaningfully better. Roche’s VP came in with a clear brief: a 1-click interface. Fewer steps, less training, faster adoption.
The research
Before any design work, I planned and ran field research across Europe and the United States, interviewing pathologists and histology technicians in their labs, mapping their actual scanning and image analysis workflows, and documenting pain points. The same research phase was replicated in Asia, by other colleagues.
One lesson from that process has stayed with me: the highest-ranking person in the room is rarely your end user. They’ll describe the problem in terms of business outcomes and organisational priorities, then hand the product to someone on their team: a technician, a researcher, a lab manager, who will actually use it every day. You have to find that person and talk to them directly. We did.
What we heard consistently was that lab environments are busy, high-stakes, and low-tolerance for anything that requires sustained attention to a screen. Technicians load slides, start a job, and move on. The interface needed to disappear.
The design
The client asked for 1-click. We proposed auto mode. 0 clicks.
The entire scanning workflow was redesigned around a single screen with an always-open settings panel. Every essential action was reachable without navigating away: configure, scan, review, save. The scanning process displayed its own progress. Quality control, including the ability to zoom into any slide section and reject images before saving, happened in the same view. Even file management, which involves large image files destined for an image server, was handled without leaving the primary interface.
The principle that emerged from user testing and shaped the whole design: “if nothing is selected, everything is selected.” It sounds wrong until you say it out loud to a pathologist who’s been using multi-step scanners for twenty years, and they nod. Tapping into people’s instincts, even when the logic seems counterintuitive, is how you design something that requires no training.
Only secondary actions like deep device settings, image server management, fire-and-forget configurations, lived behind additional clicks. And even those flows were fully designed and tested. There was no corner of the product that was left to engineering judgment.
Beyond the screen
Midway through the project, I noticed a gap in the hardware specification: there was no way for a technician to understand the state of a running job without looking at the screen. In a busy lab, that’s a problem as you’re often moving between benches, not watching a monitor.
I proposed a 6-LED panel on the device enclosure, designed to communicate job state at a glance from across the room. It wasn’t in the original spec bu it made it into the product.
Influencing the physical design of a piece of medical hardware as a UX consultant isn’t something that happens by accident. It happens because you’ve spent enough time in the labs, with the people who work in them, to understand what they actually need, and because you’ve built enough trust with the client to be heard when you bring something they didn’t ask for.
What it established
This project was midway through my career, but at a point when I completely shifted from in-house web and app design to consulting for Fortune 500 clients at Impossible; it set the template for how I approach complex, high-stakes design problems: start with research, talk to the actual users, define the workflow before designing the interface, and never stop at the brief. If there’s a better answer than the one the client asked for, find it and make the case.
For more on my approach to healthcare UX research and design, read my 2019 article Strategies for Designers Working in the Deep Healthcare Space.
The outcome
The DP200 launched and was adopted in labs across Europe, the United States, and Asia. The product is part of Roche’s Digital Pathology lineup to this day.
The success led directly to a follow-on commission: the DP600, a 240-slide version of the DP200. By then I had moved on to Design Director, but I remained involved in the early research and supported the two-person team who took the DP200 design and scaled it to the new platform. That product also shipped.
A few years later, doing field research for a different project at the Leeds Institute of Medical Research, I walked into a lab and saw a DP200 on the bench, in use. There is no better proof of work than that.
