Each Semantic Knowledge Graph contains a forward index and a complementary reverse index to represent information with intersecting meaning. These indices then represent relationships between information. As illustrated in the figure, a forward index identifies each problem and procedure for a patient, while the reverse index identifies each patient with a problem or a procedure. As a result, the level of correlation between information nodes can be scored to reveal latent relationships. This new approach coupled with the ability to build knowledge graph relationships in real-time with a proven, real-world corpus of data establishes relationships between patients, problems, and procedures directly at the time of need - not months after the patient has gone home.

KnowledgeGraphs are much more granular than the illustration above. As illustrated in the figure below, a KnowledgeGraph for a patient context contains distinct representations of patient administrative, demographics, comorbidities, past medical history, diagnosis, procedures, and events. In fact, SAM contains over 21 million conceptual relationship permutations within its correlation engine.