DIVA, a 3D virtual reality platform, improves undergraduate craniofacial trauma education

Jebrane Bouaoud, Mohamed El Beheiry, Eve Jablon, Thomas Schouman, Chloé Bertolus, Arnaud Picard, Jean‐Baptiste Masson, Roman Hossein KhonsariJournal of Stomatology Oral and Maxillofacial Surgery 2020 · DOI

TL;DR

This paper introduces DIVA, a novel 3D virtual reality (VR) platform designed to enhance the education of undergraduate medical students in craniofacial trauma. The study demonstrates that DIVA significantly improves students' understanding and visualization of complex craniofacial anatomy and injuries, with high reported satisfaction and accuracy in anatomical landmark placement.

What Problem Does This Paper Address?

Teaching craniofacial trauma is notoriously challenging due to the intricate anatomy of the head and neck. Traditional methods, relying on textbooks and 2D CT scans, often lead to misunderstandings and diagnostic pitfalls. While 3D CT scans offer an improvement, visualizing and interacting with these complex structures in a way that truly aids understanding, especially for complex trauma cases, remains a hurdle. The authors identified a gap in educational tools that leverage immersive technologies like VR to specifically address the complexities of craniofacial trauma education for medical students.

Methods

The study employed a two-pronged approach to evaluate DIVA:

  1. Educational Efficacy and User Experience: 50 undergraduate medical students were tasked with diagnosing and planning treatments for three distinct craniofacial trauma cases using DIVA. Following this, they completed an 8-item questionnaire assessing their satisfaction, perceived benefit, ease of use, and tolerance of the VR system.
  2. Anatomical Precision: Four postgraduate students were asked to identify and place six specific anatomical landmarks on 3D renderings of the same trauma cases. This was performed using both standard CT-scan-based 3D reconstructions (via Avizo software) and DIVA's VR interface. Geometric morphometrics, specifically Procrustes analysis and permutation testing, were used to compare the accuracy and variability of landmark placement between the two methods.

Datasets

The study utilized anonymized CT-scan data from three craniofacial trauma cases (jaw fracture, naso-orbital-ethmoid complex fracture, and Le Fort III fracture). These cases were presented to both undergraduate and postgraduate students. No new benchmark datasets were introduced.

Key Insights and Ideas

  • VR as a Game-Changer for Complex Anatomy: The study strongly suggests that immersive VR environments like DIVA can overcome the limitations of 2D representations for complex anatomical structures, making them more accessible and understandable for medical trainees.
  • Beyond Visualization: Enhanced Understanding: DIVA not only allows for better visualization but also appears to facilitate a deeper comprehension of craniofacial trauma, enabling students to identify lesions that might be missed on standard 2D or even conventional 3D renderings.
  • High User Acceptance and Tolerance: The platform achieved very high satisfaction rates and was generally well-tolerated, indicating its potential for widespread adoption in medical education without significant user discomfort.
  • Precision Meets Immersion: The study successfully demonstrated that DIVA's VR environment provides a level of anatomical precision comparable to traditional 3D reconstruction methods, making it a reliable tool for learning and practicing anatomical landmark identification.

Key Results

  • High Satisfaction: 92% of undergraduate medical students reported a high degree of satisfaction with DIVA.
  • Ease of Use and Tolerance: 88% found DIVA easy and intuitive to use, and 86% reported it was well-tolerated.
  • Improved Lesion Detection: 82% of students reported visualizing lesions with DIVA that were not seen on usual 2D and 3D renderings.
  • Strong Recommendation: Almost all students (98%) recommended DIVA for educational purposes in traumatology.
  • Accurate Landmark Placement: Geometric morphometrics showed no significant difference in landmark placement accuracy between DIVA and standard 3D CT-scan visualizations. Mean 3D errors for landmark positioning were consistently below 1 millimeter (ranging from 0.37 to 0.84 mm).

Limitations and Future Work

The authors acknowledge that the study involved a limited number of trauma cases and a specific cohort of undergraduate and postgraduate students. Future work could involve expanding the range of trauma cases, testing DIVA with a larger and more diverse student population, and conducting longitudinal studies to assess the long-term impact of VR training on diagnostic skills and clinical outcomes. The authors also suggest that DIVA could be valuable for exploring other complex anatomical areas beyond trauma, such as craniofacial malformations and head and neck cancers, by integrating MRI data.

Summary updated

BibTeX
@article{bouaoud2020diva,
  author = {Jebrane Bouaoud and Mohamed El Beheiry and Eve Jablon and Thomas Schouman and Chloé Bertolus and Arnaud Picard and Jean‐Baptiste Masson and Roman Hossein Khonsari},
  title = {{DIVA, a 3D virtual reality platform, improves undergraduate craniofacial trauma education}},
  journal = {Journal of Stomatology Oral and Maxillofacial Surgery},
  year = {2020},
  doi = {10.1016/j.jormas.2020.09.009},
}
Figure