What If You Could See What I See? Using AI to Elevate Medical Device Design

What If You Could See What I See? Using AI to Elevate Medical Device Design

At ClariMed, our experts bring both professional expertise and personal insights to medical device human factors. In this thought leadership piece, Hajni Salazar-Velekey, our Senior Human Factors Consultant shares how personal experience sparked fresh perspectives in accessibility testing approaches.

A Personal Glimpse Into Vision Impairment

"White silk overlay." "Dirty and blurry windscreen." And still having to drive.

That's how I've tried to describe what I see with my late-onset corneal haze. But words alone never really captured it, until I turned to ChatGPT. I asked it to manipulate an image to show others how I see through my eyes.

It didn't change anything about how I approach my clinical care, but it did something else: it helped the people around me understand my reality just a little better.

Beyond Empathy: Could AI Help Us See Through the Eyes of Others?

This experience made me wonder: if AI can help me show what I see, could it also help researchers and designers better understand the lived experiences of people with visual impairments?

Women, in particular, are more likely to live with chronic conditions and spend more of their lives in poor health, making vision challenges just one piece of a larger puzzle.

My eye condition didn't come from a chronic illness, but dealing with it gave me a glimpse into just how disruptive vision issues can be. And that made me think more deeply about those navigating both long-term health conditions and unreliable vision—especially when using medical devices that require clarity, focus, and precision.

Chronic Conditions and the Fluctuating Vision

Now picture this: Someone managing a condition like multiple sclerosis might experience blurred vision, double vision, or washed-out colors—sometimes triggered by heat or fatigue. Another, living with Type 1 diabetes, may see things shift when blood sugar dips. On some days, reading a screen or lining up an injector is no problem. On others, it's like threading a needle in a fog.

These symptoms come and go. They shift with time, environment, and energy. And when devices require precise actions—such as dosing, aligning, confirming—it can become a real barrier to confident use.

AI-Powered Simulation: Beyond Current Methods

Now imagine if AI could simulate those moments. What if we could "see" through their eyes just as I did with mine—but tailored to conditions like MS or Type 1 diabetes?

AI could offer something different—perhaps even more adaptive—than existing AR or VR simulation tools. While current platforms provide valuable fixed experiences of certain impairments, AI has the potential to go further: creating dynamic, personalized simulations that reflect how symptoms shift with time, environment, or disease state. It could help generate "vision journey maps" that capture daily fluctuations, or flag accessibility challenges earlier in device design.

Why It Matters for Human Factors?

When used in the early stages of human factors work or exploratory research, tools like these could bring us closer to creating devices that aren't just safe and compliant—but intuitive, inclusive and truly user-centered.

Could simulating real-world vision challenges help bring your device closer to truly user-centered design?

It’s a question worth exploring, and one we’re incorporating into our human factors approach.

Curious How This Could Apply to Your Product?

Our team at ClariMed continually explores innovative approaches to better understand diverse user experiences and needs.

Whether you're addressing vision challenges, cognitive considerations, or other user experience factors, we can help you identify potential usability barriers earlier in your design process and develop solutions that work for all users.

Contact our team to discuss how our evolving approaches to user-centered design can strengthen your development process and help create more accessible, inclusive medical devices.