Notwithstanding the popularity of machine learning in natural language processing (NLP), rule-based systems have their advantages: distinctive transparency, ease of incorporating external knowledge, and demanding fewer annotations. However, processing efficiency and rule complexity remain major obstacles for adopting rule-based NLP solutions in large clinical data analyses. This talk will introduce a new rule processing engine that allows fast rule execution and structured rule construction for clinical NLP that helps to reduce the rule complexity.
Understanding how to identify the social determinants of health from electronic health records (EHRs) could provide important insights to understand health or disease outcomes. We developed a methodology to capture 2 rare and severe social determinants of health, homelessness and adverse childhood experiences (ACEs), from a large EHR repository.
This presentation will describe the process and results of a recently completed formal practice analysis of Clinical Informatics conducted by the American Medical Informatics Association in collaboration with the American Board of Preventive Medicine and with the support of the American Board of Pathology. The aim of the practice analysis was to develop a comprehensive and current description of what Clinical Informatics Subspecialty physician diplomates do and what they need to know.
This talk will cover our recent work on developing deep learning algorithms with applications to biomedical narrative text. The common theme of these studies aims at building models that improve prediction accuracy by exploring and combining relational information in text.
Increasingly, vast amounts of health-related data are being generated by a diverse array of sources, such as EHRs, medical claims, product and disease registries, laboratory test results and even consumer mobile devices.
Enormous potential exists to improve oral health services throughout the world by using information and communication technologies, such as teledentistry, to expand access to primary, secondary and tertiary care. Comparison of teledentistry procedures with standard clinical procedures can demonstrate the relative effectiveness and cost of each approach. However, due to insufficient evidence, it is unclear how these strategies compare for improving and maintaining oral health, quality of life, and reducing healthcare costs.
Routinely collected patient electronic medical record (EMR) data are approaching the genomic scale in volume and complexity and is increasingly recognized as a valuable resource for clinical research to answer questions for broader populations than would have ever been possible with a specialized research environment.
This is a webinar on applications of NLP to clinical psychology. It is novel and has been primarily tackled in open domain NLP. This webinar brings together open domain NLP with clinical NLP and is expected to unify two otherwise fairly separate communities.
The aim of pharmacovigilance is to identify and describe adverse events related to the use of medicines and to support wise therapeutic decisions. By nature, pharmacovigilance focuses on the unexpected and relies on effective methods for data-driven discovery. In this talk I will highlight initiatives seeking to improve our ability to do exploratory analysis in observational medical data, using statistical pattern discovery, latent class cluster analysis and natural language processing.