What mistake taught you the biggest lesson when trying to get healthcare VR products into clinicians’ hands?
Skip Rizzo: I did not realize early on that “budget” would be such a factor. I thought if we had the science to show this can do it better than the existing method, people would be clamoring to get it. In 2003 we built a virtual classroom for assessing attention in children and everyone loved it in focus groups. Then we asked if they would spend $10,000 on a primitive headset, computer, and software. People backed off and said what they had was sufficient. Since then, hospital directors ask, how will this reduce costs and generate revenue, that is the bottom line in health.
Why is healthcare VR still not mainstream, even with the science and the technology?
Skip Rizzo: What we do not have, and why it is not mainstream, is clinician adoption on a large scale. Clinicians often rely on what they learned in graduate school. If they did not learn it there, maybe they take continuing education, but this can feel like a bridge too far. They are anxious or not familiar, and market studies show younger clinicians are more likely to adopt than old school clinicians. We need curriculums brought to graduate schools across healthcare, psychiatrists, psychologists, social workers, occupational therapy, physical therapy, and teach VR, simulation technology, and AI early on, so it is not foreign and they grow up with it.
What advice would you give to studios and teams building experiences for the healthcare VR sector?
Skip Rizzo: Look at the very front end and the very back end. On the front end, make sure what you are building is something the clinical population wants, needs, and would be willing to use. The only way is co-design and user centered design, working with targeted clinical populations early so you do not build something from the ivory tower that nobody will use. On the back end, ask how you will fund it and who will pay. In the United States it is a business, so you need to know if it is paid by insurance, out of pocket, or clinicians because it improves outcomes and credibility.
Where do you see AI supporting healthcare VR right now, with concrete examples?
Skip Rizzo: I am an AI optimist, but not blind to implementation challenges. We have written ethical and professional guidelines for patients facing AI. What has changed now is content generation for immersive healthcare and better measurements. There is a company doing exposure therapy where you talk into the headset, describe a trauma scenario, and in about 30 seconds it pops up a spherical image of that location and setting. It is still primitive and fails a lot, but it is marching forward. AI also helps software development, coding, and big data analytics, so we can quantify behavior, performance, engagement, and interaction in these worlds, not just build the worlds.
You mentioned measurement, can you give a specific example of what AI enables inside VR assessments?
Skip Rizzo: Take the virtual classroom. We used to measure total head movement, reaction time, variability, plus performance. With AI we can be more precise and tell how many times a child missed a target on the whiteboard when they were looking out the window at a school bus or people walking by, versus looking directly at the target. Those are two different attention errors, distractibility versus loss of focus. We can parse that second by second, do complex analyses in two minutes, and in a standalone headset send performance metrics to the cloud and spit back a fully fleshed out report within five minutes. That real time data can feed training and adjust challenge millisecond by millisecond.
Virtual simulations are now populated by AI characters we can talk to. How do you assess these systems so they give safe, appropriate answers to patients?
Skip Rizzo: You start from the beginning with what will populate the large language model. Are you going to take stuff randomly off the internet? Maybe for formulating language, but not for content. We built a portal to upload specific, well vetted information approved by the Veterans Association. You still monitor for hallucinations, but the system should pull from vetted content, not Reddit. Then you test, first with a less clinically oriented group. What is also important is that we are not doing therapy. As soon as we detect a red flag, we stop and push them to a live provider, with a one touch hotline or clinician contact.
What kind of collaboration is actually needed to unlock the next breakthrough and wider adoption?
Skip Rizzo: Success comes from an interdisciplinary team. In the optimal case you have clinical experts, software engineers, graphic artists, and people who understand human factors and design. Then you need people who understand business, administration, and healthcare business specifically, plus legal advice to minimize liability and marketing. That is nine different areas, and it is usually two or three people trying to do all of it, which is why a lot of startups fail. The companies I see succeed already have a marketplace, existing customers, reputation, and the backend sales, marketing, and productization in place, then they add immersive tech as a product line.
Can you share an example of a company that had the setup to make immersive rehab products feasible?
Skip Rizzo: A company that illustrates this is Bioness. They are in rehabilitation and have been around for about 30 years developing rehab tools, safety harnesses, large flat screen touch panels for upper extremity rehab, and so on. They already have a reputation and a market, around 400 solid customers across the United States. Now they want to take what they have been developing on flat screens for the last 10 to 15 years and turn it into immersive technology that expands their product line. Slipping immersive tech into an existing product line is more feasible because the sales, support, and relationships are already there.
