Designing the next 15,000 procedures.
With 15,000+ Galaxy procedures performed across 60+ clinical sites, Galaxy II had to evolve a proven platform without disrupting the workflows physicians already trusted.
As imaging choices and automation increased, my focus was maintaining procedural awareness: making it immediately clear what the system had done, what it was about to do, and where physician judgment was still required.
The Context
Galaxy II is Noah Medical’s robotic bronchoscopy platform, used by physicians to steer a robotic scope deep into the lungs and biopsy small, hard-to-reach nodules under live X-ray imaging. Think of it like piloting a highly precise robotic instrument with a video game controller — except every interaction happens during a live procedure, where confusion can cost time, disrupt clinical judgment, or increase risk to the patient.
My Role
Senior Product Designer on a three-person design team, owning key Galaxy II features plus system-wide motion and interaction specs, and co-owning the Galaxy Design System. I partnered closely with clinical, systems, software, and hardware teams to define, validate, and ship the experience.
Lesion Focus
Capability:
Lesion Focus removes metal artifacts from the live image so physicians can see the lung lesion more clearly. Think of it like Photoshop’s Remove tool, but used during robotic bronchoscopy.
Challenges & Constraints:
Decisions:
Solution
A user arrives on this screen and immediately selects the Lesion Focus button.
Alternative explorations:
Full-screen in-progress
Popover with text "Lesion focus in-progress…"
A user arrives on this screen, performs other tasks, then selects the Lesion Focus button.
Capability:
The Galaxy robot can’t tell which way the C-arm is positioned, so the physician needs to confirm its orientation before the system can proceed safely.
Example of C-arm Orientation
Current Workflow:
Challenges & Constraints:
Decisions:
Updated Workflow:
Alternative explorations:
We considered adding the warning to existing on-screen instructions, but testing showed experienced physicians often skip them. Clinical engineering recommended a more explicit interruption instead.
Design opportunity:
C-arm Orientation detection – Galaxy would have the capability to detect the orientation of the C-arm at any point of the procedure. This would remove unnecessary modal dialogs and allow the user to continue through the workflow uninterrupted.
Solution
Capability:
Galaxy II lets physicians choose between two C-arm sweep modes — HD-DT and CBCT — during the procedure. Think of it like switching camera modes depending on what you need to see: one mode may provide a faster, focused view, while the other captures a more complete 3D image.
Sweep Comparison
Challenges & Constraints:
Both modes need to feel like prominent Galaxy II features.
Each mode has different instructions, timing, and clinical implications, so confusing them could affect what happens next in the procedure.
Physicians need to understand the selection at a glance, without stopping to read.
Live View & Secondary view remained the same size per clinical requirements.
Decisions:
Designed two distinct yet visually consistent icons for each sweep mode based on the physical sweep direction and angle.
Shared design language with current C-arm Orientation selection.
Defined clear interaction states:
Disabled
Selectable
Selected
Updated on-screen instructions dynamically based on the physician’s selection, reinforcing which mode was active.
Outcome (verified):
Validated in formative usability studies with KOL physicians and nurses using a structured testing approach.
Solution
Alternative explorations:
Other components:
Toggle
Switcher/Segmented button
Design opportunity:
Automatic sweep detection – Remove the requirement for users to select a sweep type. Allow them to perform a sweep and the C-arm will make automatic adjustments as they go.
"Being able to continuously understand the scope-to-lesion relationship gives me confidence before and during sampling."
Gautam Balakrishnan, MD, FCCP | Mosaic Life Care
Takeaways
We were a small design team of three, which meant staying close to the work and making decisions quickly across a wide range of product areas.
The strongest solutions came from close collaboration with software, clinical engineering, systems engineering, and software quality assurance. Understanding what was a true technical constraint, a clinical risk, or simply a difficult implementation problem helped us make better design tradeoffs.
Not every ideal experience was technically possible within the release timeline. In those moments, UX became more than interface design. We used flows, states, and interaction models to make system logic visible, align teams, and move decisions forward.