Work
The Oculus Kiosk
The Oculus Kiosk
How did we streamline the check-in process at ophthalmology clinics for the visually impaired?
DISCLAIMER: All Oculus Kiosk concepts, research, and drawings are owned by Camille Dorset (MS, Applied Health Informatics at Fordham University). The project was developed collaboratively with John Chelsom (Seven Informatics Ltd.) and myself as a contract UX Designer.
Overview
TYPE OF WORK
Product Design, End-to-End UX Process, Service Design, Content Design, Research Paper
TIMELINE
July 2023 — April 2025
Originally scoped as a 3-month contract, I was retained to continue iterative design and research through the publication of the paper in April 2025.
TEAM
1x Lead UX Designer (me)
1x Research Project Lead / Developer
1x Subject Matter Expert (MOA)
LOCATION
Victoria, Canada 🇨🇦
Oxford, UK 🇬🇧
New York City, USA 🇺🇸
TOOLS
Figma / FigJam
Zoom
KEY WORDS
Product Design, Content Design, Digital Health, Kiosk / Mobile UX, Accessibility (a11y), Research Paper
HISTORY LESSON
The Evolution of Medical Records: From Clay to Code
Origins: The earliest health records came from ancient Mesopotamia where clay tablets or ṭuppu (𒁾𒉺) were used to record patients' health details. Physicians or asû (𒀀𒋢𒌑) and healers or āšipu (𒀀𒋢𒁉𒌑) would use the sakikkû, or a diagnostic medical manual, to understand how to treat their patient. The sakikkû (𒊕𒄭𒅕) would serve as their clinical decision support system at the time.
Today: In today's digital world, health records come in the shape of electronic health records (EHRs). But another problem arises from this new medium: how can we make these health records easy-to-use for a broad range of patients and interoperable across different clinics?
Defining the problem
The Oculus Kiosk is a research-driven initiative focused on improving both workflow efficiency and patient experience within ophthalmology clinics. Through our investigation, we identified two critical dimensions of the problem: accessibility challenges affecting patients and operational inefficiencies impacting medical office assistants (MOAs).
THE PATIENT PERSPECTIVE
👨 Patient Problem:
Healthcare is inaccessible—even ophthalmology
Read the anecdote below from the MOA I worked with:
— • —
This week, a patient came into our clinic and tried to complete our new patient forms. When they realized they couldn’t fill them out, they became overwhelmed and left in tears. The patient did not have fingers.
— Medical Office Assistant
— • —
This story was a prime example showcasing various bottlenecks in the current process including:
🎯👈
Motor Barriers
Patients who had a physical impairment often struggled with the manual paperwork due to struggles with writing or holding the pen and paper.
🔍
Hard-to-See and Use
With over 30% of Americans affected by visual impairment, many ophthalmology patients struggle with small, dense text on paper forms.
🤯
High Cognitive Load
Due to various factors including medical jargon in long symptom lists, text-heavy forms, and confusing legal disclaimers, the current forms are hard to understand.
THE PROVIDER PERSPECTIVE
👩⚕️ Provider Problem:
Manual data entry causes bottlenecks
Providers faced their own set of problems with paper-based intake forms ranging from difficult multi-tasking, inconsistent triaging, and slower check-in times.
✍️
Manual Data Entry
Paper-based intake forms create significant time overhead: patients take longer to complete written forms, while providers must spend additional time clarifying and correcting entries.
⏱️
Workflow Bottlenecks
Paper forms can make check-in take longer and slow down the clinic, especially when many patients arrive at once. This results in long queues at check-in and cognitive overload for providers.
📋
Triage Friction
Triage priority can become subjective, as providers interpret urgency differently. Due to this variability, this leads to inconsistent decisions and slower intake, as MOAs must cross-check multiple forms.
MY Objective
From Concept Sketches to an Accessible, Patient-First Prototype
Seven Informatics Ltd., a research consultancy, recently explored integrating cityEHR into ophthalmology triage. A student researcher, also an SME and former ophthalmology MOA, created proof-of-concept sketches based on clinic care pathways.
My task was to refine these sketches into an accessible and UX-driven prototype that would accommodate ophthalmology patients of all backgrounds based on the current problems faced by patient and provider.
Original concept drawings of the Oculus Kiosk created by Camille Dorset
Original concept drawings of the Oculus Kiosk created by Camille Dorset
Personas: Who are we designing for?
Based on the interviews and research, we determined two primary personas divided into patient and provider perspectives:
👴🏽
Juan
Patient
A 65-year-old patient who has a condition of cataracts and also carpal tunnel (unrelated to his visit)
Is visiting the clinic to get a routine check-up surrounding his cataracts and see if it is progressing
👩⚕️
Emily
Medical Office Assistant (MOA)
A 28-year-old medical office assistant (MOA) working at an ophthalmology clinic facilitating patient check-in, triage, administrative tasks, and clinical support
Her clinic has 20-50 patients come on a daily basis
Research & Validation
METHOD
Field Studies:
What's working? What's broken?
Getting out into the 'field': What excited me most about this project was designing for a kiosk—an interface I had never worked on before. To understand best practices, I conducted field research on self-service kiosks in the real world, including fast food chains like McDonald’s and hotel check-in systems.
Findings: After exploring self-service kiosks in real-world settings, I identified common design patterns that could inform the Oculus Kiosk. Key takeaways on navigation, touch target sizes, and element placement were summarized to guide the UI design.
METHOD
Accessibility Audit:
Designing through a different lens
Low Vision Simulator: I used a low-vision simulator in Figma as a quick way to evaluate design decisions without the need for testing participants. This helped me assess touch target sizing and typography across conditions such as blurred vision, hemianopia, and central or peripheral vision loss.
Blurred Vision
Retinal Detachment
Central Vision Loss
Peripheral Vision Loss
WCAG Contrast Checker: I also conducted contrast testing to ensure color pairings met accessibility standards and remained legible under these visual impairment conditions.
METHOD
Usability Testing:
Understanding what the users see and feel
I recruited 5 subjects for usability testing based on Nielsen's Law of Small Numbers. There were three main flows they were tasked to complete:
Checking into a scheduled appointment.
Checking into a walk-in appointment as a new patient.
Checking into a walk-in appointment as a returning patient.
Participants were recruited based on the most common eye conditions in the general population, primarily refractive errors: myopia (~27%), hyperopia (~10–20%), astigmatism (~33%), and presbyopia (~100% over age 40). The distribution of eye conditions among test participants is shown below.
FINDINGS
Insights:
Understanding what the users see and feel
Here are key findings from the various research sessions.
😶🌫️
Low Contrast
The first design iteration had low contrast and failed WCAG compliance checks while using the contrast checker.
👓
Symbolism Aids Clarity
From usability testing, users often said the icons helped them identify and recognize symptoms quicker.
💳
Auto-Populating Data
Experienced U.S. healthcare users noted they usually show their insurance card and suggested auto-filling their details from it.
The Solution
Self-Serve Kiosk:
A Medium with Alternatives
A self-serve kiosk reduces data entry for clinic staff while providing patients with a flexible, easy-to-use interface.
Due to Figma limitations, the demos were recorded using an iPad as the prototype device. However, the final specifications accounted for both iPad and kiosk deployments to accommodate varying clinic budgets.
Due to Figma limitations, the demos were recorded using an iPad as the prototype device. However, the final specifications accounted for both iPad and kiosk deployments to accommodate varying clinic budgets.
Design Decisions
IMPROVEMENT #1
🔎 Large Target Sizes:
Bigger Buttons, Better 'Fitts'
Motor Accessibility:
Inspired by an SME’s anecdote of a patient without fingers, the Oculus Kiosk prioritizes larger touch targets for accessibility. Digitizing intake removes the struggle of holding a pencil for physically or cognitively impaired patients. With oversized buttons exceeding WCAG 2.1’s 44x44 px standard, even those with limited dexterity can navigate the interface with ease.
IMPROVEMENT #2
🖼️ Intuitive Iconography:
An Icon Is Worth a Thousand Words
Cognitive Accessibility (Understandability / Comprehension Support):
An SME from a New York City ophthalmology clinic emphasized the need for clear visuals to simplify complex medical jargon. Large images and icons aid recognition, especially for visually impaired patients, and condition names were simplified for key medical conditions.
Usability testing showed 80% of participants used icons as helpful secondary aids to confirm their choices after reading the text. Icons were placed on a dark yellow (#FFB81C) background for contrast, avoiding red-green combinations for deuteranopia.
These icons also reduce the need for providers to clarify conditions, speeding up check-in, lowering MOA workload, and supporting consistent triage.
Improvement #3
🌗 High Contrast UI:
Eye-Friendly, User-Friendly
Visual Accessibility:
To enhance accessibility, I tested the Oculus Kiosk’s color palette with a color-blind peer and incorporated their feedback. I also referenced WCAG 2.1 contrast guidelines to ensure high-contrast colors that accommodate all users, including those with visual impairments like color blindness.
Improvement #4
🔤 Readability:
Type That Talks
Visual & Cognitive Accessibility:
To address the readability of the text and copy on the screen, we explored a combination of typography, font choices, weights, and concise copy.
📜
Sans-Serif
I used two sans-serif fonts—Cal Sans for titles and buttons, and Manrope for labels and body text—to enhance readability by eliminating extra strokes that serif fonts would include.
↕️
Text Size
Larger, bolded text with wider line spacing improves glanceability and lexical decision-making. Following APH ConnectCenter guidelines, text is at least 18pt (24px) for low-vision users.
🔤
Sentence Case
Using sentence case improves legibility by creating distinct word shapes, aiding those with visual impairments. Unlike title case, which appears block-like, sentence case enhances readability.
IMPROVEMENT #5
👓 Screen Reader-Friendly:
Reading Aloud for Your Ears Only
Subtle design choices were key to accessibility in the Oculus Kiosk, including logical UI order, working with the developer for clear backend labeling of elements, and alt text for all non-text visuals. Designing the kiosk to have various options of interacting with the patient allows for less time spent by MOAs helping patients fill out their intake forms.
Improvement #6
📺 Ergonomics:
Bridging Software & Hardware
Physical & Motor Accessibility:
The kiosk’s button-based interaction improves accessibility over handwritten input, especially for patients with limited dexterity. 40% of participants preferred scanning their health insurance card instead of manual entry, supporting its value for physically impaired users.
Implementing this feature requires a card scanner, clear onboarding, and a confirmation step before submission. While a full kiosk was out of scope, the concept was tested on an iPad/tablet setup, with a separate card scanner required.
Scanning your Health Insurance ID Card: The instructional graphics should bridge the gap between the software and the hardware. They are self-explanatory and provide users with system feedback and progress of their card scan. There are 3 distinct phases with 2 edge case for statuses:
Enter your card
Scanning your card
Remove your card
Scan success
Card error
Retrieving Your Queue Number: After triage, patients receive a printed queue number. An estimated wait time is included to set expectations and help reduce anxiety. The queue number is generated based on the Oculus Kiosk's triaging algorithm, saving MOAs and providers a lot of time on triaging and administrative tasks.
Results & Next Steps
RESULTS
Impact
♿️
WCAG Compliance
Accommodating various accessibility types through design decisions. Thorough adoption of WCAG 2.1 standards for the Oculus Kiosk.
👍
Improved UX
Through usability tests with 5 visually-impaired participants, we achieved a SUS score of 87.5; a 17% improvement from a SUS score of 75 from previous iterations.
📑
Published Paper
Co-authored a research paper on ontology-based kiosks for ophthalmology clinics published by Springer Nature in the HIMS 2024 conference proceedings.
Sprint II - IV Prototype
Sprint V Prototype
Published Conference Paper!
VIEW PAPER
I co-authored “A Health Records Kiosk Using an Ontology-Based Information Architecture” (J. Chelsom, C. Dorset, W. Lee, 2025), published by Springer Nature in the conference proceedings volume Health Informatics and Medical Systems and Biomedical Engineering (HIMS 2024). The paper explores how ontology-driven information architecture can improve accessibility, navigation clarity, and cognitive load in public health record systems, contributing practical HCI guidance for inclusive kiosk design.
Future
Next Steps
Develop the Oculus Kiosk
The Oculus Kiosk is currently being developed and it will soon be prepared for deployment to real-world clinics. The Oculus Kiosk uses the cityEHR created by Seven Informatics Limited as a backbone for its back-end.
Deploy to a Real-World Clinic
Once the Oculus Kiosk is fully developed, we have plans of piloting it in either the New York clinic our SME works at or in Oxford, close to where Seven Informatics Limited operates. Though it was originally designed for an American healthcare system, depending on where we decide to ship, we will make slight adjustments to the workflow.
Pilot Testing
Once we have implemented and deployed the Oculus Kiosk to a real-world clinic, we plan on doing a second round of testing with both patients along with care providers and MOAs and continue to iterate and improve the user experience.
The mobile version is currently under 🚧 construction 🚧
Meanwhile, please feel free to enjoy my portfolio on desktop.
Contact
Let's work together! Get in touch through one of these links:
© William Lee 2024
2025
William Lee