Achieving automated narrative text interpretation using phrases in the electronic medical record

¹Laboratory of Computer Science Massachusetts General Hospital, Boston, USA.

Proceedings : a Conference of the American Medical Informatics Association. AMIA Fall Symposium

|January 1, 1996

Summary

Clinicians use repetitive phrases in medical records, often linked to specific concepts. Automating interpretation of this narrative text is now possible using the MEDPhrase application and Unified Medical Language System (UMLS) concepts.

Area of Science:

Medical Informatics
Natural Language Processing

Background:

Clinicians frequently employ stereotypic phrases within electronic medical records.
These phrases often represent underlying semantic concepts specific to individual practitioners.

Purpose of the Study:

To formalize the association between stereotypic clinical phrases and semantic concepts.
To enable automated interpretation of narrative text in electronic medical records.

Main Methods:

Analysis of the COSTAR medical record database.
Utilizing the MEDPhrase application to link phrases with Unified Medical Language System (UMLS) concepts.

Main Results:

Demonstrated that stereotypic phrases can be reliably mapped to underlying semantic concepts.
Successfully automated the interpretation of narrative clinical text.

Conclusions:

Formalizing phrase-concept associations enhances the utility of electronic medical records.
The MEDPhrase application offers a scalable solution for clinical text interpretation.

Related Experiment Videos

Achieving automated narrative text interpretation using phrases in the electronic medical record

Related Concept Videos

Male-male aggression is unlikely to stabilize a poison frog polymorphism.

The Harvard-MIT-NEMC Research Training Program in Medical Informatics.

Diabetes and the risk of infections with immunomodulator therapy in inflammatory bowel diseases.

Prenatal antidepressant exposure is associated with risk for attention-deficit hyperactivity disorder but not autism spectrum disorder in a large health system.

Higher plasma vitamin D is associated with reduced risk of Clostridium difficile infection in patients with inflammatory bowel diseases.

Psychiatric co-morbidity is associated with increased risk of surgery in Crohn's disease.

Identification of findings suspicious for breast cancer based on natural language processing of mammogram reports.

Searching for information on the Internet using the UMLS and Medical World Search.

Text structures in medical text processing: empirical evidence and a text understanding prototype.

A natural language parsing system for encoding admitting diagnoses.

Meeting clinician information needs by integrating access to the medical record and knowledge resources via the Web.

Electronic forms: benefits drawbacks of a World Wide Web-based approach to data entry.

Large Language Models as Clinical Support Tools in Drug Information Services: Performance Comparison With Pharmacists.

ChatGPT-4o in delirium recognition: a pilot study.

A Methodological Approach to Assessing Closed-Loop Communication in the Operating Room.

Data Extraction From Oncology Imaging Reports by Large Language Models: A Comparative Accuracy Study.

Assessing the clinical applicability of large language models for microwave ablation guidance.

Consistency evaluation protocol: A reproducible framework for assessing large language model output repeatability.

Related Experiment Videos

Achieving automated narrative text interpretation using phrases in the electronic medical record

Area of Science:

Background:

Purpose of the Study:

Main Methods:

Main Results:

Conclusions:

Related Concept Videos

Related Articles