Sequence labeling for extraction of medical events and their attributes from unstructured text in Electronic Health Record (EHR) notes is a key step towards semantic understanding of EHRs. It has important applications in health informatics including pharmacovigilance and drug surveillance. The state of the art supervised machine learning models in this domain are based on Conditional Random Fields (CRFs) with features calculated from fixed context windows. In this application, we explored recurrent neural network frameworks and show that they significantly out-performed the CRF models.
Download full-text PDF |
Source |
|---|---|
| http://www.ncbi.nlm.nih.gov/pmc/articles/PMC5119627 | PMC |
| http://dx.doi.org/10.18653/v1/n16-1056 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Mental Health Utilization Among Transgender Veterans.
JAMA Netw Open
January 2025
VA Center for Health Information and Communication, US Department of Veterans Affairs, Veterans Health Administration, Health Systems Research CIN 13-416, Richard L. Roudebush VA Medical Center, Indianapolis, Indiana.
Importance: Compared with cisgender (CG) individuals, transgender and gender-diverse (TGD) individuals experience substantial social and economic disparities that can result in adverse mental health consequences. It is critical to understand potential barriers to care and to address the causes of the disparities in the future.
Objective: To characterize mental health care utilization among TGD veterans with depression.
Addressing Inadequate Documentation of Unilateral Do-Not-Resuscitate Orders.
JAMA
January 2025
Department of Pulmonary and Critical Care, University of Chicago, Chicago, Illinois.
Enhancing emergency care for older persons: the role and impact of the electronic Frailty Index.
Geroscience
January 2025
Department of Emergency Medicine, IRCCS Humanitas Research Hospital, Rozzano, Milan, Italy.
As the elderly population expands, enhancing emergency department (ED) care by assessing frailty becomes increasingly vital. To address this, we developed a novel electronic Frailty Index (eFI) from ED health records, specifically designed to assess frailty and predict hospitalization, in-hospital mortality, ICU admissions, and 30-day ED readmissions. This retrospective, single-center study included patients 65 years old or older who presented to the ED of IRCCS Humanitas Research Hospital in Milan, Italy, between January 2015 and December 2019.
View Article and Find Full Text PDFEarly Results of Implementing Rapid Methadone Titration for Hospitalized Patients: A Case Series.
J Gen Intern Med
January 2025
University of California, San Francisco, CA, USA.
Objectives: With the increase in illicit fentanyl use in the USA, hospitals face challenges managing opioid withdrawal and opioid use disorder (OUD). To improve opioid withdrawal and OUD treatment among hospitalized patients with daily fentanyl use, we developed a rapid methadone titration (RMT) protocol. We describe development, implementation, and outcomes during the first 12 weeks.
View Article and Find Full Text PDFFirst year of in-house forensic neuropathology consultations in Helsinki, Finland.
Int J Legal Med
January 2025
Forensic Medicine Unit, Finnish Institute for Health and Welfare, P.O. Box 30, Helsinki, FIN-00271, Finland.
In July 2023, an in-house forensic neuropathology consultation pilot was established at the Helsinki office of the Forensic Medicine Unit, Finnish Institute for Health and Welfare. This offered an alternative to the previous practice of full outsourcing to a hospital neuropathology department. This paper aims to introduce the first year experiences of the pilot.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!