Akademik Veri Yönetim Sistemi
the IX Biennial Meeting of the Society for functional near-infrared spectroscopy (fNIRS2026), Macau, Çin, 16 - 19 Ekim 2026, ss.1-5, (Özet Bildiri)
Abstract: Virtual Reality (VR) has become a widely used tool in surgical training due to its flexibility and reproducibility. However, the neural mechanisms underlying VR-based training, particularly neurovascular coupling, remain poorly understood. In this study, we investigated the longitudinal effects of VR training on skill acquisition and mental workload, and compared them with traditional physical simulation using a multimodal fNIRS–EEG approach. A longitudinal design was employed across multiple training sessions. Preliminary results revealed significant differences, particularly in temporal dynamics and delayed neural responses. These findings highlight the potential of multimodal biomarkers for improving surgical skill assessment and training optimization. Introduction: VR simulators have become increasingly integrated into surgical education. Despite their widespread adoption, the effectiveness of VR training and the mechanisms through which it enhances surgical performance remain subjects of ongoing debate. In particular, there is a lack of longitudinal studies employing multimodal approaches to investigate mental workload, motor learning, and skill retention in VRbased training. Moreover, the underlying neural and cognitive mechanisms remain poorly understood. The primary aim of this study is to investigate the neurovascular mechanisms underlying VR-based surgical training using a multimodal approach combining fNIRS and EEG, and to compare these findings with traditional physical simulation training. The secondary aim is to predict behavioral performance and skill metrics using induvial and combined neurophysiological features. Methods: Twenty-eight surgical residents and medical students were enrolled in this study (data collection is ongoing, and recruitment for the VR group is still in progress). Participants were assigned to one of two groups: a Virtual Reality (VR) training group or a physical simulation training group. The VR group performed laparoscopic tasks (peg transfer and string pass) in a virtual environment, while the physical group performed the same tasks using a conventional box trainer. Participants attended five sessions: an initial baseline session, three intermediate self-training sessions, and a final session. During the baseline and final sessions, hemodynamic and neural activity were simultaneously recorded using an 18-channel fNIRS system and an 8-channel EEG system. The experimental protocol consisted of a 2-minute resting baseline followed by 4 minutes of task performance for each task. During the intermediate self-training sessions, participants practiced the assigned tasks for 10 minutes per task without neurophysiological recording. In addition to neurophysiological data, behavioral performance metrics including NASA-TLX workload scores, completion time, number of errors, and aptitude test results were recorded as secondary measures. Results: Preliminary analysis indicates significant differences between baseline and final sessions in behavioral and performance metrics, particularly within the physical training group. Improvements were observed in error rates. In addition, multimodal analysis revealed temporal delays and altered coupling patterns from baseline to the final stage, suggesting changes in neurovascular dynamics over the course of training. Ongoing analyses will further compare these effects between VR and physical training groups. Conclusion: These preliminary findings suggest that neurovascular biomarkers derived from multimodal fNIRS–EEG data may provide insights into the neural mechanisms underlying surgical skill acquisition. This approach has the potential to enhance objective assessment and certification processes, particularly in VRbased training environments. However, several limitations should be considered, including differences in haptic feedback and levels of immersion between VR and physical simulations. These factors will be further discussed in the present study