What is it about?
Artificial intelligence (AI) has become an engine of modern medicine, helping clinicians analyze medical images, predict risks, and tailor treatments by finding patterns in vast amounts of health data. At the same time, extended reality (XR), including virtual, augmented, and mixed reality, is gaining popularity both in medicine and beyond because it reshapes how people experience and interact with digital information layered onto the real world. The real excitement lies in combining the two: extended reality provides an interactive space directly within a patient’s field of view, while AI personalizes and adapts what happens inside it, responding to each patient in real time. Together, they point toward a future of healthcare that is not only more technologically advanced but also more human-centered, bringing personalized, engaging care closer to patients’ everyday lives. In this systematic review, we analyzed 64 studies reporting interconnected AI-XR applications for patient-facing health and well-being. The research shows that these applications are particularly promising in three areas: rehabilitation (helping patients recover motor skills after stroke or injury), mental health (treating anxiety, phobias, and PTSD through immersive therapy), and developmental care (supporting children with autism or ADHD with real-time nudges). AI enhances XR experiences by personalizing difficulty levels, tracking patient progress, and creating responsive virtual environments that adapt in real time. While outcomes are encouraging with patients showing improved engagement and better clinical results, we also identified significant hurdles: the technology is still maturing, most studies involve small patient groups, and ethical questions around data privacy in immersive environments need attention.
Featured Image
Photo by Neuro Equilibrium on Unsplash
Why is it important?
This review arrives at a critical moment. XR headsets are becoming mainstream consumer devices, and generative AI is advancing rapidly. However, healthcare applications lag behind workspace productivity and entertainment use cases. Our work provides the first comprehensive map of how these technologies can be combined to benefit patients, highlighting what works and what doesn't. For healthcare and computational researchers, the review aims to provide four generalizable paradigms for successfully integrating both powerful technologies. In addition, it highlights substantial roadblocks for further adoption, including the need for larger validation studies and ethical considerations that must be addressed before these tools can scale. The timing matters: as AI-powered XR tools enter clinical settings, having a clear evidence base and a robust ethical foundation will help ensure patient safety and meaningful outcomes, rather than relying on blind trust in hyped technology.
Perspectives
I started this review because I believe in the increasing adoption of extended reality glasses in our daily lives. While famous disruptors, such as Meta CEO Mark Zuckerberg, predict smart glasses will eventually replace smartphones as our primary computation platform, I advocate for research-backed, sustainable, and ethical adoption across domains, including healthcare. In this review, we seek to highlight the exciting potential of emerging technologies. Imagine a future where we can get affordable clinical and therapeutic support in the comfort of our own homes. Imagine how we can receive health guidance in daily situations like grocery shopping or physical workouts, fully on demand and tailored to us. We're not there yet, but this review highlights the path and the roadblocks moving forward.
Tim Schwirtlich
Northwestern University
Read the Original
This page is a summary of: Synergy of Artificial Intelligence and Extended Reality in Patient-focused Health and Well-Being Applications – A Systematic Review, ACM Transactions on Computing for Healthcare, February 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3793544.
You can read the full text:
Contributors
The following have contributed to this page







