Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jul 1, 2026

Assessment and Communication for People with Disorders of Consciousness
07:37

Assessment and Communication for People with Disorders of Consciousness

Published on: August 1, 2017

Automated Prediction of Glasgow Coma Scale Scores From Unstructured Electronic Health Records Using Natural Language

Marta Fernandes1, Niels Turley2, Haoqi Sun2

  • 1Department of Neurology, Massachusetts General Hospital, 55 Fruit St, Boston, MA, 02114, United States, 1 8573319160.

Journal of Medical Internet Research
|June 29, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

CSF Testing for Neuroinvasive West Nile Virus and Measures to Improve Guideline Adherence.

Neurology. Clinical practice·2026
Same author

Delirium and Increased Risk of Developing Dementia: An Emulated Target Trial Analysis.

medRxiv : the preprint server for health sciences·2026
Same author

Gabapentinoid Polypharmacy Among Medicare Beneficiaries During the Poststroke Recovery Period.

Health services research·2026
Same author

Epidemiology of Traumatic Brain Injury in Medicare Beneficiaries over the Last Decade: A Retrospective Observational Study.

Health science reports·2026
Same author

Late-Onset Seizures: Etiology and Demographics in US Tertiary Care Epilepsy Centers.

Neurology·2026
Same author

Post-Discharge Anti-Seizure Medication Use Improves Post-Stroke Survival: <i>An Emulated Target Trial in Older Adults</i>.

medRxiv : the preprint server for health sciences·2026
This summary is machine-generated.

This study developed a natural language processing model to accurately predict Glasgow Coma Scale (GCS) scores from electronic health records. This tool enhances critical care research by enabling large-scale neurological assessment phenotyping.

Area of Science:

  • Critical Care Medicine
  • Biomedical Informatics
  • Natural Language Processing

Background:

  • Electronic health records (EHRs) are valuable for critical care research but face challenges in extracting key clinical variables like patient consciousness levels.
  • Accurate assessment of neurological status is crucial for quality improvement and comparative effectiveness research.

Purpose of the Study:

  • To develop and validate a natural language processing (NLP) model for predicting Glasgow Coma Scale (GCS) scores using daily EHR notes.
  • To enable automated, large-scale phenotyping of neurological assessments in critical care settings.

Main Methods:

  • A pooled ordinal regression model (ordinalNet) and a pooled linear model were trained on EHR data from Mass General Brigham and MIMIC-III databases.
  • Models predicted GCS impairment levels (severe, moderate, mild) and continuous scores using daily notes, age, sex, and admission type.
Keywords:
Glasgow Coma Scaleelectronic health recordsmachine learningnatural language processingphenotyping

More Related Videos

Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
07:31

Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack

Published on: May 15, 2020

IntelliSleepScorer, a Software Package with a Graphic User Interface for Mice Automated Sleep Stage Scoring
04:54

IntelliSleepScorer, a Software Package with a Graphic User Interface for Mice Automated Sleep Stage Scoring

Published on: November 8, 2024

Related Experiment Videos

Last Updated: Jul 1, 2026

Assessment and Communication for People with Disorders of Consciousness
07:37

Assessment and Communication for People with Disorders of Consciousness

Published on: August 1, 2017

Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
07:31

Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack

Published on: May 15, 2020

IntelliSleepScorer, a Software Package with a Graphic User Interface for Mice Automated Sleep Stage Scoring
04:54

IntelliSleepScorer, a Software Package with a Graphic User Interface for Mice Automated Sleep Stage Scoring

Published on: November 8, 2024

  • Performance was evaluated using AUROC, AUPRC, RMSE, and Pearson correlation coefficients.

    • Main Results:

    • The pooled ordinalNet model achieved an AUROC of 0.96 and AUPRC of 0.77 for predicting GCS impairment levels.
    • The pooled linear model demonstrated a root mean square error of 2.30 and a Pearson correlation of 0.76 for continuous GCS scores.
    • Model predictions correlated with specific clinical terms indicating patient alertness and interventions.

    Conclusions:

    • Developed NLP models accurately predict GCS scores from unstructured EHR data.
    • These models can significantly support large-scale phenotyping of neurological assessments in critical care research.
    • The findings facilitate enhanced data abstraction for quality improvement and comparative effectiveness studies.