Assessing the Feasibility of an AI-Enhanced VR Program Delivering Personalized Mind-Body Therapy for Depression - PROJECT SUMMARY This proposal aims to conduct a pilot randomized controlled trial (RCT) to test the feasibility and acceptability of XAIA (eXtended-reality Artificial Intelligence Ally)—a novel, adaptive intervention combining virtual reality (VR) with artificial intelligence (AI) to deliver personalized mind-body therapy—vs. sham VR control in patients with depression. Major depressive disorder is common and consequential, affecting over 8% of adults in the US annually and causing profound morbidity and mortality. Mind-body therapy is an evidence-based intervention for depression, yet its real-world impact is limited by challenges in access due to workforce shortages and variable geographic distribution of mental health providers. While digital, on-demand mind-body therapy programs offer a promising solution to bridging the gap between supply and demand of mental health professionals, meta- analyses show that digital mental health apps are less effective and have lower adherence vs. face-to-face therapy. These findings may be due to therapists' ability to foster engagement, active participation, and treatment retention, in contrast to self-administered digital mental health programs where non-adherence and early discontinuation reduce its clinical impact. To address these challenges, our multidisciplinary team developed a digital mind-body therapy intervention designed to enhance clinical outcomes by improving engagement, adherence, and completion of a sufficient therapy dose. In Phase I (Design) of the ORBIT Model, we created XAIA—a VR program where users interact with and are guided by an AI conversational avatar that engages in adaptive, natural language conversations with the patient and personalizes the mind-body therapy experience. This, in turn, enhances the relevance of the mind-body therapy content, improves engagement, ensures a sufficient dose of therapy, and more effectively transfers these skills in users. Consequently, this may lead to improved cognitive flexibility and emotion regulation skills, which ultimately leads to improvements in depression and quality of life. In an initial study published in Nature Digital Medicine, patients reported after one-time use of the AI/VR mind-body therapy app that it was a promising and safe tool for their depression. We are now in Phase II (Preliminary Testing) of the ORBIT Model and propose to conduct a pilot RCT comparing the AI/VR mind-body therapy program vs. sham VR control to achieve the following aims: Aim 1—Establish the acceptability of using an 8-week AI/VR mind-body therapy program among patients with depression; Aim 2—Assess the feasibility to recruit, randomize, and retain participants with depression in both the intervention and control arms in an 8-week pilot RCT. If this pilot study shows evidence that the AI/VR mind-body therapy program is feasible and acceptable among patients with depression, it will provide a strong platform to undertake a definitive Phase II efficacy trial in a future R01. Given the high prevalence and significant societal impact of depression, along with a shortage of clinicians, developing and validating innovative, on-demand, complementary approaches—such as XAIA— that enable people to self-manage their symptoms via mind-body approaches can fulfill a critical unmet need.
