Abhiram Kanneganti

Opportunities for medical students to gain practical experience in vaginal births are increasingly constrained by shortened clinical rotations, patient reluctance, and the unpredictable nature of labour. To alleviate clinicians' instructional burden and enhance trainees' learning efficiency, we introduce a mixed reality (MR) system for childbirth training that combines virtual guidance with tactile manikin interaction, thereby preserving authentic haptic feedback while enabling independent practice without continuous on-site expert supervision. The system extends the passthrough capability of commercial head-mounted displays (HMDs) by spatially calibrating an external RGB-D camera, allowing real-time visual integration of physical training objects. Building on this capability, we implement a coarse-to-fine localization pipeline that first aligns the maternal manikin with fiducial markers to define a delivery region and then registers the pre-scanned neonatal head within this area. This process enables spatially accurate overlay of virtual guiding hands near the manikin, allowing trainees to follow expert trajectories reinforced by haptic interaction. Experimental evaluations demonstrate that the system achieves accurate and stable manikin localization on a standalone headset, ensuring practical deployment without external computing resources. A large-scale user study involving 83 fourth-year medical students was subsequently conducted to compare MR-based and virtual reality (VR)-based childbirth training. Four senior obstetricians independently assessed performance using standardized criteria. Results showed that MR training achieved significantly higher scores in delivery, post-delivery, and overall task performance, and was consistently preferred by trainees over VR training. Although validated only in the context of obstetric delivery, the system demonstrates strong potential for broader manikin-based procedural training and other healthcare education scenarios.

Opportunities for medical students to gain practical experience in vaginal births are increasingly constrained by shortened clinical rotations, patient reluctance, and the unpredictable nature of labour. To alleviate clinicians' instructional burden and enhance trainees' learning efficiency, we introduce a mixed reality (MR) system for childbirth training that combines virtual guidance with tactile manikin interaction, thereby preserving authentic haptic feedback while enabling independent practice without continuous on-site expert supervision. The system extends the passthrough capability of commercial head-mounted displays (HMDs) by spatially calibrating an external RGB-D camera, allowing real-time visual integration of physical training objects. Building on this capability, we implement a coarse-to-fine localization pipeline that first aligns the maternal manikin with fiducial markers to define a delivery region and then registers the pre-scanned neonatal head within this area. This process enables spatially accurate overlay of virtual guiding hands near the manikin, allowing trainees to follow expert trajectories reinforced by haptic interaction. Experimental evaluations demonstrate that the system achieves accurate and stable manikin localization on a standalone headset, ensuring practical deployment without external computing resources. A large-scale user study involving 83 fourth-year medical students was subsequently conducted to compare MR-based and virtual reality (VR)-based childbirth training. Four senior obstetricians independently assessed performance using standardized criteria. Results showed that MR training achieved significantly higher scores in delivery, post-delivery, and overall task performance, and was consistently preferred by trainees over VR training.

Background/Objectives: Our study evaluates the performance of the CDO1 and CELF4 methylation assay of cervical scrapings in diagnosing endometrial cancer. Methods: Patients referred for abnormal uterine bleeding, suspected endometrial pathology on imaging, endometrial hyperplasia, or cancer, who were indicated for endometrial sampling +/− hysteroscopy +/− hysterectomy, were recruited prospectively from the gynecology clinic of Prince of Wales Hospital, Hong Kong, from January 2023 to November 2024. Methylation assay was performed on the cervical scrapings of the participants, with the results compared to endometrium histology obtained on the same day. Results: Methylation assay results of 675 patients were compared to 629 outpatient endometrial assessments and 46 hysterectomy histology. The assay yields an accuracy of 97.3%, sensitivity of 84.1%, specificity of 98.8%, PPV of 89.2%, and NPV of 98.2%. The AUC is 0.92 with a Kappa coefficient of 0.85. The false-negative rate was 0.8% among the 629 endometrial assessment cases. Endometrial thickness, abnormal cervical cytology, and endometrial hyperplasia significantly affected the accuracy. Among seven false-positive cases, five had endometrial hyperplasia and two had cervical intraepithelial neoplasia. Eventually, three patients were diagnosed with endometrial cancer and one with cervical cancer 1 to 4 months later. Conclusions: Methylation assay of cervical scrapings is highly accurate in diagnosing endometrial cancer with a high sensitivity, specificity, and low false-negative rate. It can act as a triage to reduce invasive endometrial assessment. All false-positive cases were related to neoplastic processes in the genital tract, indicating that it may be useful for detecting cancer early, before histological change is evident.

Simulation-based training is a vital tool in obstetrics and gynaecology education, addressing challenges such as limited clinical exposure during short rotation periods and gender-related sensitivities. This need is further underscored by the fact that non-obstetricians and non-midwives often serve as first responders in out-of-hospital and acute care settings where vaginal birth (VB) is imminent. Traditional mannequin-based simulations depend on in-person tutors with demands on time and physical space. In the absence of realistic clinical scenarios, they serve no better than low-fidelity task trainers. Capitalising on recent advancements in affordable virtual reality (VR) head-mounted display (HMD) technology, we developed and iteratively refined a VR-based childbirth training system between 2020 and 2025. Initial user studies demonstrated strong knowledge retention although participants expressed a preference for in-person mannequin-based simulations due to the lack of tactile feedback in VR. To address this limitation, we introduced a mixed reality (MR)-enhanced childbirth training system in 2024. This innovative approach integrates high-precision alignment of virtual elements with physical mannequins and introduces a tactile component to enhance skill acquisition. A pilot study involving 35 Year 4 undergraduate students yielded positive feedback with participants praising the system’s heightened realism and opportunities for peer-to-peer interaction. While both VR and MR have distinct advantages and limitations, the affordability and versatility of modern HMD make VR/MR-based training modules highly scalable and adaptable to diverse scenarios, surpassing the single-task limitations of traditional mannequins. This promises to improve training efficiency and expand access to VB training for a broader range of healthcare professionals.

Objective: To compare knowledge acquisition and domain‐specific feedback between virtual reality (VR) and mannequin‐based simulation for teaching medical students normal vaginal delivery (NVD). Methods: We conducted a cluster randomized crossover study comparing a novel in‐house‐developed NVD VR simulator utilizing an Oculus® Quest 2 HMD with the PROMPT Flex® mannequin. We randomly assigned medical student clinical groups to either VR or mannequin simulation. Participants took an 11‐item knowledge quiz before and after their first simulation before crossing over to the alternate simulation. Participants took a six‐domain, 14‐item feedback questionnaire evaluating both simulations. The primary outcome was percentage improvement in knowledge scores, while the secondary outcome was adjusted mean difference in mean feedback scores between each simulation. Results: Twenty clinical groups comprising 111 medical students participated. There was no difference in either arm for baseline characteristics and pre‐simulation knowledge scores. Medical students undergoing VR simulation had significantly higher overall post‐simulation mean percentage correct scores (79.0% vs. 54.0%; P < 0.001). The mannequin simulator had significantly higher feedback scores (61.0 vs. 58.3; adjusted mean difference 2.45 [95% confidence interval 0.95–3.95]; P = 0.002) and in five of six domains. A total of 40.5% of respondents preferred the mannequin simulation, 4.5% preferred the VR simulation, and 42.5% preferred a mixture of the two. Conclusions: Our novel VR NVD simulator was superior to traditional mannequin simulation in content delivery. As more respondents preferred a mixed approach over any single method, VR simulation could complement legacy mannequin simulation methods by serving as a primer or a distance‐learning tool and help increase skill birth attendant training.

Advancements in Virtual Reality (VR) technology have opened new frontiers in medical education, igniting interest among medical educators to incorporate it into mainstream curriculum, complementing traditional training modalities such as manikin training. Despite numerous VR simulators on the market, their uptake in medical education remains limited. This paper explores the acceptability and educational effectiveness of VR in the context of vaginal childbirth delivery training, with the simulator providing a walkthrough for the second and third stages of labour, contrasting it with established manikin-based methods. We conducted a large-scale empirical study with 117 medical students, revealing a significant 24.9% improvement in knowledge scores when using VR as compared to manikin. However, VR received significantly lower self-reported feasibility scores in Confidence, Usability, Enjoyment, Feedback and Presence, indicating low acceptance. The study provides critical insights into the relationship between technological innovation and educational impact, guiding future integration of VR into medical training curricula.