Project coordinators :
- Jocelyne Troccaz (gmcao,TIMC)
- Olivier Palombi (imagine, LJK)
- Violaine Cahouet (saiga, GIPSA)
- Laurence Nigay (iihm, LIG)
Abstract:
In this Persyval team we propose to develop the « Living Book of Anatomy » also called LBA to see the working/acting body using Augmented Reality techniques. The idea is to capture a user action (limb movement, speech sound or orofacial movement) and to visualize information about how is the muscles that are responsible for it, superimposed onto the user image. Such tool could be used to facilitate learning anatomy for medicine students and to provide a interactive way for general public to discover human anatomy in action.
The assumption underlying this work is that experiencing the anatomical knowledge with his/her own body, or embodiment, could make its acquisition easier. We aim at evaluating this hypothesis as well as the interactive designed techniques on real medicine students. Another hypothesis is that the use of Augmented Reality could make the embodiment easier by a closer connection between models, knowledge and reality. Advances in numerical sciences, the development of web services, applications and the spread of smart phones and tablets in general and more specifically in the clinical world, resulted in quite a large collection of accessible data and applications describing anatomy or other medical knowledge but as far as we know, none of them bridges the gap between the user’s own body and this knowledge.
Developing the LBA requires the resolution of several scientific or technical issues: (1) capturing the user action, (2) mapping the model to the user body and conversely, (3) rendering anatomical and functional information onto the user body (4) designing augmented reality interaction techniques. We will see in the detailed description of the work that these issues can be solved with different levels of complexity. Underlying the system, a suitable knowledge base must be built, using OBDA (Ontoly-Based Database Access), including medical knowledge combining the description of anatomical structures and organs, together with their functions and behaviors, also including biomechanical models, medical images, and so on. Usable and efficient interactive techniques should make the knowledge accessible, with several levels of refinement, for several types of users with various levels of skills. One challenge is also to make the LBA system able to render such information in real-time or interactive time.
This Persyval team initially connects four laboratories on top of 3 of the 4 actions of the labex [Persyval]. Of course, the action “Authoring Augmented Reality (AAR)” is the most relevant to the proposed topic. Two of the three WPs are directly concerned: (WP2.1) real-time capture and simulation and (WP2.3) natural interaction with the augmented world. The action “Modeling and simulating for understanding the real world (SIM)” is obviously concerned by the proposed work: (WP3.1) multi-scale, multi-physics modeling is necessary for knowledge base construction and exploration. Also (WP3.4) inverse problems has some relevance as it will be discussed in the presentation of the intended work. Finally, the progress of “Advanced Data Mining (ADM)” may also benefit to this work. To summarize, the laboratories intend primarily to contribute to the AAR advances and will benefit from SIM and ADM ones. TIMC-IMAG, LJK, GIPSA-lab and LIG are the initial actors of the Persyval LBA team. They are strongly connected to potential users of the LBA system: school of Medicine, LADAF and clinicians of the CHU. The breadth of the related topics obviously potentially opens to other laboratories of Persyval-Lab and to laboratories related to human sciences (didactics, psychology). We will promote these intra-Persyval and extra-Persyval connections by organizing a bi-monthly seminar related to the topics of the LBA.
During the 3,5 years of the team lifespan, the laboratories will contribute to the LBA through their own research activity and by collaboratively supervising research fellows (PhD and post-doctoral students) on identified bottlenecks. Also, a deliverable of the team will be the actual development of two “chapters” of the LBA and their evaluation to confirm that the underlying assumptions are true.
