Research organizations have to process a large amount of documents in plain text, often from patients’ electronic health records, which have been long recognized as valuable but unstructured sources of information. Saving time and effort by applying machine learned models to big datasets could greatly reduce the amount of time and labor required to conduct a literature review or make sense of an unwieldy quantity of data. Healthcare natural language processing (NLP) applications are on track to become an industry worth $2.7 billion in 2020, thanks in no small part to a general push to digitize all medical records.
Technology Description
Our technology is an interactive natural language processing platform that enables domain experts such as clinicians and clinical experts to make use of NLP techniques using a point-and-click interface over the cloud. After an initial round of training, it combines novel text-visualizations to help its users make sense of NLP results, revise models and understand changes between revisions. It allows the users to make any necessary corrections to computed results, thus forming a feedback loop and helping improve the accuracy of the models. This tool can be used for relevant applications in the medical domain for both clinical research purposes, either as a standalone research tool or integrated with existing systems. These techniques could also be generalized to other fields, such as legal management systems.Advantages
Reduces time, energy, and labor required to review and evaluate prohibitively large data sets or large volumes of clinical texts
Allows users to make corrections to computed results, creating a feedback loop to improve model accuracy
Easy-to-use point-and-click interface’Applications
Extracting important information from unstructured EMR data
Medical research with large datasets
Literature reviews and preparatory research
Clinical applications
Legal management systemsStage of Development
Working prototype designed to help colonoscopy researchers evaluate quality metrics from large volumes of clinical text.IP Status
Copyright (software)Relevant publications
Trivedi, G., Pham, P., Chapman, W., Hwa, R., Wiebe, J., Hochheiser, H. NLPReViz: an interactive tool for natural language processing on clinical text. Journal of the American Medical Informatics Association, Volume 25, Issue 1, January 2018, Pages 81–87. DOI: 10.1093/jamia/ocx070External links
NLP Platform DemoInnovators
Guarav Trivedi, PhD
Former PhD student, University of Pittsburgh School of Computing and Information Science
Dr. Trivedi is former graduate student in Intelligent Systems Program at University of Pittsburgh. His research interests are in intelligent interfaces and human-computer interaction in healthcare. During his time at Pitt, he worked on designing interactive methods for natural language processing for clinical records.Education
PhD, Intelligent Systems, Artificial Intelligence, University of Pittsburgh
B.Eng, Information Technology, National Institute of Technology in KarnatakaPublications
Jaromír Šavelka, Gaurav Trivedi, and Kevin Ashley. 2015. Applying an Interactive Machine Learning Approach to Statutory Analysis. In Proceedings of the 28th International Conference on Legal Knowledge and Information Systems.
Braga, Portugal – Awarded the Best Student Paper
Gaurav Trivedi. 2015. Clinical Text Analysis Using Interactive Natural Language Processing. In Proceedings of the 20th International Conference on Intelligent User Interfaces Companion (IUI Companion ’15). ACM. New York, NY.
Gaurav Trivedi, Phuong Pham, Wendy Chapman, Rebecca Hwa, Janyce Wiebe, Harry Hochheiser. 2015. An Interactive Tool for Natural Language Processing on Clinical Text. Presented at 4th Workshop on Visual Text Analytics (IUI TextVis 2015), Atlanta, GA.
Gaurav Trivedi, Phuong Pham, Wendy Chapman, Rebecca Hwa, Janyce Wiebe, and Harry Hochheiser. 2015. Bridging the Natural Language Processing Gap: An Interactive Clinical Text Review Tool. Poster presented at the 2015 AMIA Summit on Clinical Research Informatics.Harry Hochheiser, PhD
Associate Professor, Department of Biomedical Informatics
Associate Professor, Intelligent Systems Program
Director, Biomedical Informatics Training Program, University of Pittsburgh School of Medicine
Dr. Hochheiser has published more than 25 peer-reviewed journals, conference papers, and two book chapters. He is currently working on the development of highly-interactive, user-centered systems for finding and exploring biomedical datasets.Education
Postdoctoral, National Institute of Aging
PhD, University of Maryland
MS, Massachusetts Institute of Technology
BS, Massachusetts Institute of TechnologyPublications
Grizzle AJ, Hines LE, Malone DC, Kravchenko O, Hochheiser H, Boyce RD. Testing the face validity and inter-rater agreement of a simple approach to drug-drug interaction evidence assessment. J Biomed Inform. 2020 Jan;101:103355. doi: 10.1016/j.jbi.2019.103355. Epub 2019 Dec 12.
Warner JL, Dymshyts D, Reich CG, Gurley MJ, Hochheiser H, Moldwin ZH, Belenkaya R, Williams AE, Yang PC. HemOnc: A New Standard Vocabulary for Chemotherapy Regimen Representation in the OMOP Common Data Model. J Biomed Inform. 2019 Jun 22. doi: 10.1016/j.jbi.2019.103239
Trivedi G, Hong C, Dadashzadeh ER, Handzel RM, Hochheiser H, Visweswaran S. Identifying incidental findings from radiology reports of trauma patients: An evaluation of automated feature representation methods. International Journal of Medical Informatics. 2019 Sept 129; 81-7. DOI: 10.1016/j.ijmedinf.2019.05.021
King A, Cooper GF, Clermont G, Hochheiser H, Hauskrecht M, Sittig DF, Visweswaran S. Using machine learning to selectively highlight patient information. Journal of Biomedical Informatics. 2019; Oct 29;100:103327. DOI: 1016/j.jbi.2019.103329.