219 Design · Medical Device

Photocoagulator GUI

Laser eye treatment, refined. From seed stage to production, a team of three built a GUI to introduce a new way to interact with an archaic optic medical device.

Role: UX/UI Designer
Team: Product Manager, SW Engineer, Designer
Timeframe: ~1.5 years, weekly syncs

Project

Our team faced a few core challenges: selecting an appropriate GUI color scheme for the device, building features for technical practices newly introduced to the optometry community, and testing voice-activated user interfaces. Most user research was conducted by the client team, while our team concentrated on product development and design, iterating and testing prototypes with a diverse set of users before launch.

Process

Research & analysis

Market research and competitive analysis of similar systems guided initial design brainstorms — collecting information about what users were used to, what we could improve upon, and what features were used during treatment. Comparing systems side by side let us define important primary functions and distinguish them from secondary, nice-to-have features cluttering the interface. These insights shaped persona backgrounds, workflows, and user journeys.

Competitive research board comparing necessary GUI controls across existing laser systems — Ellex, Topcon, and Lumenis tablets — with a table of functions, units, and display behavior highlighted in teal.

Studying currently marketed technology gave us a better understanding of how our system should function and address user needs. Click to expand.

User interviews & observation

We visited users and observed their current habits on competitor systems to identify pain points and needs that couldn't have been surfaced remotely, giving us a broader perspective on how environmental factors change the way users interact with a system. A shared repository for interview data — observations, videos, notes — eliminated assumptions and increased efficiency during brainstorming; clients could focus on new insights and felt more inclined to contribute to design reviews.

Shared research repository: photos of the existing Iridex laser console in an exam room, handwritten notes titled 'Questions for surgeon,' and reference images of the physical laser hardware and controls.

A shared repository of exam-room photos, device references, and interview notes kept the whole team working from the same observations. Click to expand.

Workflows & personas

For many system features, we had to understand the clinical workflow and surgical tasks that would define how the user interacted with our product. Building workflows and mapping journeys framed our goals for each task and guided design decisions. Three main personas anchored our discussions throughout — the goal was to address both the familiar veteran user and the likelihood a new user would become a returning one.

User workflow flowchart mapping preset and micropulse tasks: load, view, delete, edit, and create a preset, creating a preset with micropulse, and loading a preset from the treatment page, each broken into step-by-step task boxes.

Mapping every preset and micropulse task end to end, from creation through loading a treatment.

Set up Preset workflow diagram with screenshots for each step — start on preset page, add new treatment, set power/duration/interval, toggle micropulse, select duty cycle, set power factor, save, and load — annotated with how three different surgeon personas approach each step.

Setting up a preset, step by step, annotated with how each persona approaches the task differently.

Set up Treatment workflow diagram with screenshots for each step — start on main page, set and titrate power/duration/interval, turn on micropulse, select duty cycle, set power factor, apply, and begin treatment — annotated with how three different surgeon personas approach each step.

Setting up a live treatment, mapped the same way to compare against the preset flow.

Sketch, iterate, wireframe, iterate

Multiple rounds of iteration uncovered a product-transition issue: no matter how flashy or new our GUI screen was, users would not adopt a new system that took too long to learn. Screen interactions that strayed too far from the original physical controls were heavily pushed back on and would render a prototype useless. Whether it was primary laser controls, voice controls, or treatment presets, the process required a steady cycle of brainstorming and scrapping ideas over nine months.

Hand-sketched low-fidelity wireframes on graph paper — doctor lists, treatment presets, power/interval/duration dials, and new-user setup screens, with marginal notes on interaction questions.

Sketched low-fidelity wireframes from brainstorming sessions and feature iterations. Click to expand.

Grid of digital GUI iterations for the Dr. Norlase treatment screen across nine months, showing shifting layouts, dial styles, and color themes from dark navy to light backgrounds.

Nine months of digital iteration — testing layouts, dial styles, and color themes before converging on a direction.

Prototype testing

From paper prototypes to software simulations, we put UI screen ideas in front of users early and throughout the process, shaping our information architecture, the fluidity of interactive elements, and the overall GUI color theme. Voice command let users stay hands-free and eyes-focused on treating the correct anatomy — testing a diverse group of users with a functional prototype helped bridge a new model of interaction with a medical device.

Screen recording of the final Photocoagulator GUI prototype, navigating the Main, Settings, and Presets tabs and adjusting power, duration, and interval dials.

The final prototype in action — power, duration, and interval controls tested end to end.

Results

Our client successfully launched the product to clinical researchers and surgeons who use the system in their clinics to optimize laser eye-surgery treatment. The touchscreen GUI gave the product an edge over many competitors already on the market, with continued testing and adjustment based on ongoing user needs.