Investigating the Benefits of Physical Models for Anatomical Education in Augmented Reality

Research Background and Issues

• Identified Problems or Challenges: Traditional anatomy education extensively uses physical models, but with the development of augmented reality (AR) technology, many systems lack the cognitive benefits provided by physical models. While AR can offer dynamic and visualized learning experiences, it cannot replicate the tactile interaction advantages of physical models. Additionally, there is limited research on combining physical models with AR for anatomy education and its impact on long-term knowledge retention.
• Significance: Anatomy education requires effective teaching tools to enhance learning outcomes, especially for conveying complex information to medical students and patients. Improved understanding of anatomical information by students and patients could lead to better treatment outcomes and self-management of diseases in both educational and clinical contexts.
• Research Motivation and Related Work: Existing studies lack in-depth exploration of the combination of AR and physical models in the field of anatomy, and there are technical challenges (e.g., real-time object tracking). This study aims to fill this knowledge gap by investigating the role of physical models in AR systems, with considerations for patient education and clinical applications.

Solution

• Proposed Solution: The authors designed a design space encompassing the design parameters for integrating AR systems with physical anatomical models and developed three educational systems: 1) Tangible AR, which combines AR with a physical spine model; 2) Virtual AR, which is entirely based on virtual holographic models; and 3) Screen-Based, which uses interactive 3D models on a desktop computer.
• Innovations: This work integrates the advantages of tangible interaction and AR technology, proposes a novel design space, demonstrates its generative capabilities, and conducts comparative experiments on knowledge acquisition, memory retention, and clinical education effectiveness.
• Implementation Steps and Key Technologies:
  1. Construct a design space defining model, visualization, and interaction dimensions (e.g., model size, virtual element positioning, physical interaction properties).
  2. Develop the three systems and design learning materials (including digital 3D models and textual descriptions).
  3. Conduct a comparative study (39 participants) to evaluate knowledge acquisition and learning experiences, and explore clinical application challenges through expert evaluations.
  4. Use real-time tracking technology (Vuforia) on Microsoft Hololens 2 to integrate virtual visual content with physical models.

Research Outcomes

• Specific Outcomes:
  1. Developed a complete design space and validated its generative capabilities for designing new AR systems.
  2. Experiments revealed no significant differences in knowledge acquisition and memory retention among the learning systems (Tangible AR, Virtual AR, Screen-Based).
  3. Qualitative feedback highlighted the advantages of physical models in learning experiences, such as improved tactile interaction intuitiveness and content realism.
  4. Clinical experts emphasized the potential benefits of combining physical models with AR for patient education, but noted time constraints and technical complexity as major barriers to practical implementation.
• Advantages Compared to Existing Solutions:
  • Compared to using AR alone, this study integrates the tactile advantages of physical models, enabling users to interact with and understand virtual content more intuitively.
  • Provides valuable feedback on learning experiences and clinical applications.
• Experimental or Evaluation Results:
  • Knowledge test scores showed no significant main effects or interaction effects.
  • Memory retention remained high after one week, challenging the applicability of the "forgetting curve" model.
  • Clinical evaluations indicated that combining physical models with AR helps patients relate to anatomical structures, but a balance between technical support and patient experience is needed.
• Limitations and Future Directions:
  1. Limitations:
     • The clinical scope was narrow, focusing only on one type of arthritis (axial spondyloarthritis, axSpA), with a limited sample of experts.
     • Patients were not included as participants, limiting the validation of long-term impacts on patient education and health outcomes.
     • The study did not employ explicit pre-test baseline assessments, restricting inferences about knowledge gains.
  2. Future Directions:
     • Expand research to other anatomical domains or disease types.
     • Explore personalized patient models and designs based on individual CT data.
     • Investigate long-term effects and the potential to enhance patient self-management behaviors.
     • Improve tracking technology for physical models to enhance audiovisual experiences and accessibility for patient users.

Conclusion

The authors integrated AR and physical models, proposing multiple unique design dimensions, including model attributes, visualization priorities, and interaction methods, to explore the potential of combining physical models in education and clinical settings. This study provides design guidelines and preliminary validation results for integrating physical models with AR in anatomy education, while identifying key challenges in related technologies and clinical integration. It offers valuable directions for creating more realistic and comprehensible learning experiences for a broader range of users.