Haptic VR Training Sharpens Dental Implant Drilling Skills and Eases Mental Load

Placing a dental implant is one of the most unforgiving procedures in oral surgery. A drill must enter the jawbone at precisely the right angle, at exactly the right depth, and stop within millimeters of vital structures such as the inferior alveolar nerve. Now a team of Turkish researchers has shown that a haptic virtual reality simulator can train residents to hit those targets faster and more accurately, while simultaneously measuring how hard their brains are working as they learn. The study, published in BMC Medical Education, offers one of the most detailed looks yet at how surgical skill and cognitive workload evolve together in a simulated implant osteotomy environment.

The research, led by İlhan Şengül and colleagues at Harran University in Şanlıurfa, together with collaborators at Van Yüzüncü Yıl University, Fırat University, and Şanlıurfa Oral and Dental Health Hospital, recruited 20 oral and maxillofacial surgery residents. Twelve were classified as novices and eight as experienced based on their residency training level. Each participant completed six implant osteotomy sessions on the same standardized virtual case, spaced roughly 48 hours apart. The design was deliberately tight: a single-center, prospective repeated-measures study in which every resident performed every session, producing complete data with no dropouts and no missing measurements.

What makes the study technically interesting is the way performance was captured. Haptic virtual reality systems do not merely display a three-dimensional surgical scene; they resist the user’s hand with computed forces, so a drill that touches virtual bone feels like a drill touching real bone. In this simulator, the osteotomy procedure was scored automatically along five objective dimensions: total procedure time, angular deviation from the planned trajectory, vertical deviation, coronal deviation, and apical deviation. Because the system records these values itself, there is no rater bias, no subjective grading, and no need for an examiner to watch the drill. Every degree and every millimeter of error is logged by the machine.

The results after six sessions were striking. Procedure time dropped by an estimated mean of 30.73 seconds from the first to the sixth session. Angular deviation, the most clinically critical measure of drilling accuracy, fell by 4.34 degrees. Coronal deviation decreased by 0.82 millimeters and apical deviation by 1.32 millimeters. Vertical deviation also appeared to fall by 0.53 millimeters in the primary statistical model, although the authors caution that this particular finding was not consistently confirmed across their sensitivity analyses and should therefore be interpreted cautiously. All other improvements survived the study’s rigorous statistical corrections, including Holm adjustment across the five primary outcomes.

The statistics behind these numbers deserve attention because the researchers used a method rarely seen in surgical education studies. Rather than simple before-and-after comparisons, they fitted linear mixed-effects models with a random participant intercept, allowing each resident’s baseline ability to be accounted for individually. Type III tests with the Kenward–Roger correction provided robust estimates of how much change could be attributed to repeated practice. This framework is powerful because it uses all six sessions of data from all 20 participants simultaneously, rather than collapsing everything into a single endpoint comparison, and it handles the correlated structure of repeated measurements within the same person.

Alongside the objective drill metrics, the team measured something far more elusive: cognitive workload, the subjective mental effort a trainee invests in a task. They used the Raw NASA Task Load Index, a shortened version of the instrument originally developed by NASA to quantify the demands placed on aircraft pilots. After the first and sixth sessions, participants rated their perceived workload in a Turkish-language administration of the questionnaire. The scores told a dramatic story: overall workload decreased by 22.18 points between the two assessments, a change with a 95 percent confidence interval of −29.42 to −14.94 points and a p value below 0.001. In plain terms, the residents reported that the same procedure that once felt overwhelmingly demanding had become dramatically easier on the mind.

But here is the twist that makes the study genuinely thought-provoking. Experienced residents did not report lower workload than novices; they reported higher. Across both assessment time points, experienced participants scored 9.47 points higher on overall perceived workload than their novice colleagues, with a 95 percent confidence interval of 2.23 to 16.71 points and a p value of 0.013. One plausible interpretation is that experienced surgeons carry a richer mental model of the procedure, tracking more anatomical variables, more potential complications, and more subtle aspects of technique, and that this heightened awareness registers as greater mental demand. Another is that novices, lacking a full appreciation of the procedure’s complexity, simply underestimate how much there is to monitor. Either way, the finding challenges the intuitive assumption that experience always means less mental strain.

Equally notable is what the study did not find. There was insufficient evidence that the rate of improvement differed between novice and experienced residents on any of the five objective performance measures, and there was also insufficient evidence that the reduction in cognitive workload differed between the two groups. In other words, within the limits of this sample, both novices and experienced residents appeared to benefit from repeated haptic virtual reality practice to a broadly similar degree. The authors report no significant session-by-experience-group interaction for any objective outcome, suggesting that the simulator’s training effect was not confined to one end of the experience spectrum.

The authors are refreshingly explicit about the boundaries of their conclusions. Because the study did not assess post-training retention or whether skills acquired in the simulator transfer to real patients, the findings should not be read as evidence of clinical competence or clinical skill transfer. The Turkish-language form of the Raw NASA-TLX used in the study had not undergone formal cross-cultural validation, which the authors acknowledge as a limitation. What the study does establish is that haptic virtual reality can serve as a standardized, complementary environment in which simulator-derived performance and perceived cognitive workload can be monitored objectively over repeated sessions, something traditional apprenticeship-based training cannot easily provide.

For surgical educators, the implications are considerable. Implant osteotomy is a procedure where millimeter-level errors can have lifelong consequences for patients, yet training opportunities on real patients are limited by ethics, cost, and patient safety. A haptic simulator that automatically records angular, coronal, and apical deviations offers a way to track a resident’s learning curve with machine-graded precision, session after session, without exposing anyone to risk. The parallel finding that self-reported mental effort falls as objective accuracy rises suggests that workload measurement could serve as a complementary signal of when a trainee has moved from conscious, effortful control to smoother, more automatic execution. As virtual reality hardware becomes cheaper and more realistic, studies like this one point toward a future in which surgical training is measured not by hours logged but by degrees of error eliminated and points of mental load released.

Subject of Research: Neuroscience and Medical Simulation

Article Title: Performance and cognitive workload during haptic virtual reality-based implant osteotomy training: a prospective repeated-measures study

Article References: Şengül, İ., Utkun, M., Dayanan, C., Orak, B., Seyyidoğlu, S. M., & Çavana, A. (2026). Performance and cognitive workload during haptic virtual reality-based implant osteotomy training: a prospective repeated-measures study. BMC Medical Education. https://doi.org/10.1186/s12909-026-10471-9

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10471-9

Keywords: haptic virtual reality, dental implant training, cognitive workload, surgical simulation, NASA-TLX, osteotomy, medical education, Performance, cognitive, workload, during, haptic