[{"data":1,"prerenderedAt":25},["ShallowReactive",2],{"news-en-dental-anxiety-vr":3},{"slug":4,"lang":5,"title":6,"summary":7,"date":8,"category":9,"tags":10,"cover":16,"order":17,"author":18,"authorTitle":18,"authorSlug":18,"people":19,"bodyLang":5,"html":24},"dental-anxiety-vr","en","arXiv | Can VR help in the most frightening seconds at the dentist? FreedomAI, HKU Faculty of Dentistry, CUHK-Shenzhen and partners propose an event-triggered system for dental anxiety","Dental anxiety rises and falls as treatment unfolds, yet most dental VR only offers continuous distraction or relaxation. The team broke a simulated wisdom-tooth extraction into 5 moments of anxiety and matched each with its own support. In a 24-person randomized controlled study, anxiety scores in the intervention group were on average 12.8 points lower than in the control group (out of 100).","2026-11-16","论文",[11,12,13,14,15],"VR","dental anxiety","human-computer interaction","emotion regulation","stomatology","/images/news/dental-anxiety-vr/cover.jpg",1.5,"",[20,21,22,23],"李健铨","王本友","王俊文","赵俊杰","\u003Cp>Many people are afraid of the dentist. Studies cited in the paper estimate that about \u003Cstrong>15.3%\u003C/strong> of adults worldwide have dental fear or anxiety. Some keep putting off their visits until a small problem grows into one that needs more complex treatment.\u003C/p>\n\u003Cp>Anxiety is not constant either. Take wisdom-tooth extraction under local anesthesia, where the patient stays awake throughout: lying down in the chair, they worry about not knowing what comes next; their heart skips when instruments come into view; the sound of the drill makes them tense up; the anesthetic has been given but they still feel pain; they think it is almost over, but the treatment goes on... Within a single visit, the &quot;frightening seconds&quot; differ, and so does the help people need.\u003C/p>\n\u003Cp>Virtual reality (VR) is already used to ease dental anxiety, but the usual approach plays scenery, music or games throughout, offering continuous distraction or relaxation. Could each frightening moment instead get support that fits it?\u003C/p>\n\u003Cp>Researchers from The Chinese University of Hong Kong, Shenzhen, the Faculty of Dentistry of The University of Hong Kong, FreedomAI, Shenzhen Stomatology Hospital (Pingshan) of Southern Medical University, Peking University, the Beijing Institute of Collaborative Innovation and other institutions offer one answer in their paper &quot;On Edge in the Dental Chair: Designing VR Support for Moments of Dental Anxiety&quot;: trigger the matching support at specific events during treatment.\u003C/p>\n\u003Cblockquote>\n\u003Cp>\u003Cstrong>Paper\u003C/strong>: On Edge in the Dental Chair: Designing VR Support for Moments of Dental Anxiety\u003Cbr>\u003Cstrong>Link\u003C/strong>: \u003Ca href=\"https://arxiv.org/abs/2609.15867\" target=\"_blank\" rel=\"noopener noreferrer\">https://arxiv.org/abs/2609.15867\u003C/a>\u003Cbr>\u003Cstrong>Format\u003C/strong>: arXiv preprint (human-computer interaction, September 2026)\u003Cbr>\u003Cstrong>Affiliations\u003C/strong>: The Chinese University of Hong Kong, Shenzhen; Faculty of Dentistry, The University of Hong Kong; Beijing Institute of Collaborative Innovation; Hunan University; Shenzhen Stomatology Hospital (Pingshan) of Southern Medical University; The Hong Kong University of Science and Technology (Guangzhou); FreedomAI; Peking University; MIT Media Lab; The Australian National University\u003C/p>\n\u003C/blockquote>\n\u003Cp>\u003Cspan class=\"news-figure\">\u003Cimg src=\"/images/news/dental-anxiety-vr/fig1-overview.jpg\" alt=\"Figure 1: Overview of the event-triggered VR support system. E1–E5 are the moments of anxiety in the simulated treatment, I1–I4 the matching supports, and I0 the pause-and-exit control available throughout (source: the paper)\" loading=\"lazy\" />\u003Cspan class=\"news-caption\">Figure 1: Overview of the event-triggered VR support system. E1–E5 are the moments of anxiety in the simulated treatment, I1–I4 the matching supports, and I0 the pause-and-exit control available throughout (source: the paper)\u003C/span>\u003C/span>\u003C/p>\n\u003Ch2>First, find out which moments are most frightening\u003C/h2>\n\u003Cp>The team began with three rounds of formative studies:\u003C/p>\n\u003Cul>\n\u003Cli>\u003Cstrong>Interviews with 36 stakeholders\u003C/strong>: 12 patients, 12 dental students and 12 clinical staff, to learn when the dental setting becomes frightening and what kind of support people want.\u003C/li>\n\u003Cli>\u003Cstrong>Dentists in the design process\u003C/strong>: 3 dentists turned the interview findings into a concrete scenario, the extraction of one impacted mandibular wisdom tooth under local anesthesia, and drafted the initial moments of anxiety and support modules.\u003C/li>\n\u003Cli>\u003Cstrong>Patients walk through it first\u003C/strong>: 12 patients experienced the simulated procedure without any intervention in a Meta Quest 3 headset and were then interviewed; the team used their feedback to refine how each moment is presented and how the moments connect.\u003C/li>\n\u003C/ul>\n\u003Cp>The result was 5 &quot;moments of anxiety&quot;, each paired with exactly one kind of support:\u003C/p>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Moment of anxiety\u003C/th>\n\u003Cth>Matching support\u003C/th>\n\u003C/tr>\n\u003C/thead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>E1 Worry about the unknown procedure\u003C/td>\n\u003Ctd>Procedural explanation: a virtual dentist explains what this step involves, how it will feel and how long it will last\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>E2 Seeing dental instruments\u003C/td>\n\u003Ctd>Visual modulation: the instruments are moderately blurred\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>E3 Hearing the drill\u003C/td>\n\u003Ctd>Auditory modulation: the drill sound is turned down\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>E4 Unexpected pain despite anesthesia\u003C/td>\n\u003Ctd>Self-regulation: a &quot;breathing light&quot; guides slower breathing\u003C/td>\n\u003C/tr>\n\u003Ctr>\n\u003Ctd>E5 Treatment unexpectedly running long\u003C/td>\n\u003Ctd>Emotional support: a virtual companion appears and offers comfort\u003C/td>\n\u003C/tr>\n\u003C/tbody>\u003C/table>\n\u003Cp>In addition, participants can raise their left hand at any time to pause and then choose to continue or exit, without waiting for the system to offer support.\u003C/p>\n\u003Cp>The team distilled four design principles: support should address the needs of the present moment; patients should always keep control and a safe way out; support should be moderate and unobtrusive, so the visual and auditory modulations only &quot;soften&quot; stimuli rather than &quot;remove&quot; them, letting people still follow the treatment; and every support and control should be easy to understand, hence an onboarding tutorial, situational explanations from the virtual dentist and clear status feedback.\u003C/p>\n\u003Cp>\u003Cspan class=\"news-figure\">\u003Cimg src=\"/images/news/dental-anxiety-vr/fig2-system.jpg\" alt=\"Figure 2: The five moments of anxiety and their supports: procedural explanation, visual blurring, lower drill volume, breathing light and virtual companion; raising the left hand pauses at any time (source: the paper)\" loading=\"lazy\" />\u003Cspan class=\"news-caption\">Figure 2: The five moments of anxiety and their supports: procedural explanation, visual blurring, lower drill volume, breathing light and virtual companion; raising the left hand pauses at any time (source: the paper)\u003C/span>\u003C/span>\u003C/p>\n\u003Ch2>A 24-person randomized controlled study: the biggest differences in the two most stressful moments\u003C/h2>\n\u003Cp>The team recruited 24 adults (who had visited a dentist in the past six months or reported dental fear and anxiety) and randomly assigned them to two groups of 12. The intervention group experienced the full system with support; the control group went through the same simulated procedure without support. Both groups could pause or exit at any time. After each moment of anxiety, participants rated their anxiety on a 0–100 visual analogue scale (VAS-A); the study also recorded heart rate, skin conductance and head orientation, followed by interviews.\u003C/p>\n\u003Cp>The results:\u003C/p>\n\u003Cul>\n\u003Cli>\u003Cstrong>Lower overall\u003C/strong>: averaged across the five moments, the intervention group&#39;s anxiety scores were \u003Cstrong>12.83 points\u003C/strong> lower than the control group&#39;s (95% confidence interval −24.42 to −1.70).\u003C/li>\n\u003Cli>\u003Cstrong>Biggest differences in the last two moments\u003C/strong>: \u003Cstrong>32.7 points\u003C/strong> lower at unexpected pain after anesthesia (E4) and \u003Cstrong>24.4 points\u003C/strong> lower when treatment ran unexpectedly long (E5), both still significant after correction for multiple comparisons; the differences at the first three moments were not significant.\u003C/li>\n\u003Cli>\u003Cstrong>Physiological signals point the same way\u003C/strong>: relative to rest, the intervention group&#39;s rise in heart rate before each rating was on average about 3 beats per minute smaller than the control group&#39;s, and skin conductance responses were also lower.\u003C/li>\n\u003Cli>\u003Cstrong>Safe\u003C/strong>: no adverse events occurred in the study, and the intervention group&#39;s VR sickness scores were also lower than the control group&#39;s (5.56 vs 11.46).\u003C/li>\n\u003C/ul>\n\u003Cp>\u003Cspan class=\"news-figure\">\u003Cimg src=\"/images/news/dental-anxiety-vr/fig3-results.png\" alt=\"Figure 3: (a) Anxiety scores at the five moments, intervention group in blue and control group in orange; (b) baseline-adjusted differences between the groups with 95% confidence intervals (source: the paper)\" loading=\"lazy\" />\u003Cspan class=\"news-caption\">Figure 3: (a) Anxiety scores at the five moments, intervention group in blue and control group in orange; (b) baseline-adjusted differences between the groups with 95% confidence intervals (source: the paper)\u003C/span>\u003C/span>\u003C/p>\n\u003Ch2>The limits of the findings\u003C/h2>\n\u003Cp>The paper is clear about how far its conclusions go:\u003C/p>\n\u003Cul>\n\u003Cli>The experiment took place in a simulated setting, so it cannot demonstrate clinical effects in real treatment.\u003C/li>\n\u003Cli>It evaluated the combined effect of the full set of supports and cannot separate the contribution of individual modules such as breathing guidance or the virtual companion.\u003C/li>\n\u003Cli>The five moments always appeared in the same order and the sample was small, so factors such as the passage of time and gradual habituation may be mixed in with the effect of the support.\u003C/li>\n\u003C/ul>\n\u003Cp>The team sees personalizing support around each patient&#39;s main worry and comparing different timings of support as worthwhile next steps, with the eventual goal of testing the system in real dental visits.\u003C/p>\n\u003Ch2>The FreedomAI team\u003C/h2>\n\u003Cp>Several FreedomAI members took part in this research: COO \u003Cstrong>Junjie Zhao\u003C/strong> is a co-first author of the paper (Faculty of Dentistry, The University of Hong Kong; Beijing Institute of Collaborative Innovation); Chief AI Scientist \u003Cstrong>Benyou Wang\u003C/strong> (The Chinese University of Hong Kong, Shenzhen) is one of the corresponding authors; Chief Medical Scientist \u003Cstrong>Junwen Wang\u003C/strong> (Faculty of Dentistry, The University of Hong Kong) is the paper&#39;s last author; and CEO \u003Cstrong>Jianquan Li\u003C/strong> took part in the research with FreedomAI as his affiliation. Professor Zuolin Jin of Shenzhen Stomatology Hospital (Pingshan) of Southern Medical University and other experts contributed to the study, and the user study was approved by the hospital&#39;s ethics committee.\u003C/p>\n\u003Cp>&quot;Letting everyone keep their natural teeth&quot; starts with people being willing and able to get to the dentist in time. From the oral-health large model DentalGPT to VR that helps patients through their most stressful moments, FreedomAI wants professional dental care to understand people better and be easier to accept.\u003C/p>\n\u003Cp>Learn more: \u003Ca href=\"/en/dentalgpt#xr\">DentalGPT and core technology\u003C/a>, \u003Ca href=\"/en/dentalvista\">DentalVista 3D oral and human anatomy\u003C/a>\u003C/p>\n",1794841746252]