The MITRE Corporation
Unlocking the Patient Record for Translational Medicine
We are seeing the confluence of three disruptive changes in healthcare and the life sciences: mandated adoption of electronic medical records, next generation genomic sequencing technologies, and advances in biomedical informatics. This confluence presents an unprecedented opportunity to “unlock” the information encoded in individual patients’ records and correlate it with genetic and environmental variables at both the patient and the population level. Unlocking the data will require automated tools to extract and integrate diverse kinds and scales of information, from patient phenotype to individual genetic variations to metagenomic data (e.g., the human microbiome). These tools will rely on information captured in curated biomedical databases as well as the knowledge from the biomedical literature. We will focus on the role of standards, including “gold standard” curated data, to facilitate data exchange across different knowledge sources and to assess the effectiveness of the tools needed to take advantage of these rapidly growing resources.
Dr. Lynette Hirschman is Director for Biomedical Informatics in the Information Technology Center at the MITRE Corporation in Bedford, MA. She received her Ph.D. from University of Pennsylvania in 1972 in Formal Linguistics and held positions at the New York University Linguistic String Project, Unisys Defense Systems, and the MIT Spoken Language Systems group before joining MITRE in 1993. Dr. Hirschman’s research has focused on human language understanding, information extraction and biomedical informatics. Since joining MITRE, her work has included development of the MiTAP system for mining newswire for disease outbreak surveillance, text mining for metadata capture for metagenomics, and her current research on secondary use of clinical data linked to translational medical applications. This research has led to the open source release of the MITRE Identification Scrubber Toolkit (MIST) for automated de-identification of unstructured clinical narrative.
Dr. Hirschman has been a leader in the area of evaluation of natural language processing systems in the biomedical area. She was involved as participant and organizer of the early Message Understanding Conference (MUC) evaluations. She is a founding organizer of BioCreative (Critical Assessment of Information Extraction for Biology), the first international challenge evaluation of text mining for the biomedical domain and is also a co-organizer for the recent i2b2 NLP Challenge Evaluations.