Extended Reality to Improve Delivery Room Care Training

Maternal and neonatal mortality in delivery rooms represents a global health challenge. Annually, approximately 300,000 women die due to delivery complications, primarily from severe bleeding, while over 800,000 newborns die due to intrapartum-related events. Among the strategies to increase the number of healthcare providers able to deliver effective perinatal emergency care, simulation-based education has emerged as a promising approach. Ranging from physical manikins to software, healthcare simulation offers the opportunity to train users in a safe and controlled environment. Despite its potential, few existing medical simulators train simultaneously all the manual, procedural and non-technical skills required in the maternal and neonatal domain. Indeed, currently available simulators mainly focus on technical or procedural skills and lack integration of non-technical abilities such as teamwork, decision-making, and stress management, unless a high-resource, high-fidelity scenario is built. This thesis presents the design, development, and evaluation of innovative Extended Reality simulation tools to support healthcare professionals in managing complex emergencies in the delivery room, with a specific focus on both neonatal and maternal care. This research aims to build systems for the simultaneous training of technical, procedural and non-technical skills through immersive technologies such as Mixed Reality and visuo-haptic simulations. A major contribution of this work is the development of RiNeo MR, a Mixed Reality simulation platform designed for newborn resuscitation. The system integrates a sensorized newborn manikin with a virtual environment. The hardware includes orientation, pressure and position sensors that monitor head alignment, ventilation, and chest compressions in real time. This data is streamed to a Unity3D application where a virtual newborn is synchronized and overlapped to the physical manikin, allowing users to perform manual actions with their hands while immersed in a realistic virtual scenario. Moreover, the platform includes real-time feedback about user performance. The usability of the simulator has been tested in both immersive and non-immersive (screen-based) configurations with 11 people with no medical expertise and 5 pediatric residents. Results indicated high levels of perceived usability, realism, and engagement. Importantly, performance feedback collected during the simulation was aligned with clinical standards, offering a robust educational tool with strong potential for real-world application. After the evaluation of the first RiNeo MR prototype, the platform was significantly enhanced to support a dual-player configuration that reflects the collaboration of neonatal emergency care. Moreover, the tracking of the physical objects was improved by developing a vision-based system using the Vuforia SDK. The system adopts a server-host-client architecture, where one computer runs the simulation as host and the second connects as a client. To minimize latency and ensure consistent synchronization, both systems are connected via Ethernet, and IP addresses are predefined to align with the data transmission schema of the embedded microcontroller in the sensorized manikin. To guide the simulation process, a Finite State Machine was introduced to encode the clinical protocol. It governs the sequence of actions based on the Neonatal Resuscitation Algorithm, dynamically adapting the scenario flow according to the user’s performance. In parallel, I designed, developed, and evaluated a visuo-haptic simulation system to train fundamental surgical skills such as incision, suture, and depth perception. The simulator integrates a haptic interface (Geomagic Touch) with a visuo-haptic model of the skin developed in SOFA, an open-source framework for physical simulations. I conducted two different studies to evaluate the setup usability and performance. The first study focused on evaluating whether repeated exposure to visuo-haptic simulations could improve manual dexterity. A virtual needle-threading task was developed and tested with 44 right-handed participants who were divided into four groups: those performing 132 repetitions with the dominant hand, those with the non-dominant hand using the visuo-haptic simulator, a group completing the training with the dominant hand on a physical simulation, and subjects who did not perform any training. All participants were tested before and after the training. Results indicated that even a limited number of repetitons significantly reduced task execution time, thus suggesting that visuo-haptic simulation might be an effective complementary tool for skill acquisition in surgical education. A second study was conducted to evaluate how aging influences performance within a visuo-haptic training environment. The study involved 39 participants divided in two age groups (20–40 and 50–70 years), each completing three visuo-haptic surgical tasks: incision, dexterity, and suturing. Performance metrics such as cut length, time, trajectory path length, depth, and error rate were calculated. Results revealed that younger participants consistently outperformed older ones across all tasks. Moreover, both age groups exhibited performance gains with practice, even though younger participants showed more substantial improvements than older subjects. Usability questionnaires revealed that user experience and perceived workload were comparable across age groups, indicating that the simulator was well-accepted and usable by both younger and older adults. Future, longitudinal and cross-sectional studies are planned to optimize the developed simulations, investigate the impact of extended practice within these environments, and identify optimal training programs. Overall, this work represents a step toward integrating innovative technologies into medical training, by demonstrating the potential of Extended Reality for preparing professionals to manage critical situations. This project shows that Extended Reality offers modular, cost-effective, and scalable solutions enabling high-fidelity simulation-based education even with limited resources, thus expanding access to high-quality medical training for students and healthcare providers.