Presented By: Robert Mackin, AU-UGA Medical Partnership
Co-Authors: Erika Mackin, MIT Lincoln Laboratory

Purpose 
Our medical curriculum combines multiple basic science and clinical threads into a case-based, small group learning program that is both spiral and organ-based. It is also a two-pass system where basic science with clinical correlations predominate in the first pass; while disease states, with supporting basic science, are the foundation for the second pass. While this structure provides an effective learning environment for students, it presents a logistical challenge for faculty coordination. We are currently mapping our curriculum using a relational database. However, the output is a list of sessions, which makes it difficult to discern relationships between sessions. As the relationships between threads and years is critical to the success of our program, we feel that visualizing elements of the curriculum in two dimensions will improve faculty coordination. 

Methods 
We loaded both sessions and temporal relationships of our first module into a graph database. Session entries included instructor and discipline or thread as properties, and relationship entries included connections to either a calendar timeline or another session. After designing an initial graph schema, the application generated a model of the module's content. A graph query language was used to extract subsets of sessions along with their pre-defined relationships. The resulting models were used to explore relationships between sessions. 

Results 
We have successfully created queries that retrieve and visualize subsets of sessions and their relationships. Using the resulting model, we can recognize where sessions from different threads share complementary concepts and how they need to be chronologically organized to facilitate student learning. 

Conclusion 
Results from trials of our graph database provide an extremely useful mechanism for visualizing elements of our curriculum that would benefit from coordination across the component threads and years. We plan to incorporate additional sessions and keywords into our database to further optimize our curriculum.