2008 Poster Sessions : Taking Historical Inventories in Nonischemic Heart failure – Etiologies and Risk Factors Trial (THIN-HEART) Heart Failure Origin Organization Task (HOOT)

Student Name : David Kao
Advisor : Russ B. Altman
Research Areas: Artificial Intelligence
Patients with nonischemic dilated cardiomyopathy (NIDCM) are often labeled ‘idiopathic’ when a specific, single cause cannot be identified. Several groups have attempted to determine the prevalence of known causes of NIDCM in patients with initially unexplained NIDCM varying results. Such studies have been difficult to perform, as there is rarely sufficient evidence for a definitive diagnosis, and criteria for diagnoses which rely on historical features may be inconsistent or not universally accepted. Furthermore, quantification of risk factors is challenging due to fragmentation of clinical data regarding individuals. Clinical data sources frequently have disparate representation schema with variable degrees of clinical relevance. Nevertheless, application of robust, consistent classification criteria is required to examine the relationship between known NIDCM risk factors on a population basis. We hypothesize that abstraction of clinical data into Unified Medical Language System Concepts will facilitate automated risk factor identification and patient phenotype description. Knowledge regarding risk factor definition and patient classification can be represented in the Web Ontology Language (OWL) and the Semantic Web Rule Language (SWRL), and codification of available clinical data and diagnostic criteria will allow more semantically complex investigation for discovery of missed diagnoses and new disease associations. Once created, these knowledge bases and querying strategies can be reused on data abstracted from any clinical data repository into this schema.

I believe that modern informatics techniques have the potential to produce the next great revolution in clinical medicine. I have spent most of my career starting in college studying the intersection of basic science, engineering, and clinical medicine. I started with an undergraduate degree in biomedical engineering, studying molecular genetics and biomedical polymer engineering before going on to medical school, both at the Johns Hopkins University. I recently completed my internal medicine residency and chief residency at Stanford University where I was introduced to the field of biomedical informatics, and I am now working on several projects at the Stanford Center for Biomedical Informatics Research involving data integration for applications of clinical care delivery, research, and patient safety monitoring.