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: May 24, 2026

Methodology for Establishing a Community-Wide Life Laboratory for Capturing Unobtrusive and Continuous Remote Activity and Health Data
11:21

Methodology for Establishing a Community-Wide Life Laboratory for Capturing Unobtrusive and Continuous Remote Activity and Health Data

Published on: July 27, 2018

Can NLP Detect Loneliness in Electronic Health Records? A Proof-of-Concept Study.

Tricia Park1, Sheida Habibi1, Jane Lowers2

  • 1Department of Biomedical Informatics, Emory University, Atlanta, GA, US.

Studies in Health Technology and Informatics
|May 23, 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

Self-Reported Tianeptine Experiences on Reddit: Natural Language Processing-Assisted Qualitative Study.

JMIR infodemiology·2026
Same author

Minimum Sample Size Requirements for Machine Learning: A Study on Diabetic Neuropathy Prediction.

Studies in health technology and informatics·2026
Same author

Extracting patient reported cannabis use and reasons for use from electronic health records: a benchmarking study of large language models.

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

Monitoring novel psychoactive substance trends on social media: Analysis of discussions and dashboard implementation.

Drug and alcohol dependence·2026
Same author

A cannabis use Reddit dataset for aspect-based sentiment analysis.

Data in brief·2026
Same author

Development and Validation of an Image-Based Deep Learning Tool for Identification of Xylazine-Associated Wounds.

Journal of addiction medicine·2026

Detecting loneliness severity in electronic health records (EHRs) using natural language processing (NLP) is challenging. Models struggled due to sparse documentation, indicating documentation, not modeling, is the main limitation for computational phenotyping.

Area of Science:

  • Medical Informatics
  • Computational Linguistics
  • Clinical Research

Background:

  • Loneliness is a significant clinical concern often under-documented in electronic health records (EHRs).
  • This under-documentation hinders secondary data use and computational phenotyping for loneliness.
  • Accurate identification of loneliness severity is crucial for patient care and research.

Purpose of the Study:

  • To evaluate the efficacy of natural language processing (NLP) methods in detecting and classifying loneliness severity from clinical notes.
  • To assess the performance of transformer models and large language model-based summarization for this task.

Main Methods:

  • Retrieved clinical notes from patients with documented loneliness survey results (mild, moderate, severe).
  • Applied an expert-expanded lexicon and fine-tuned transformer models (RoBERTa, ClinicalBERT, Longformer) for classification.
Keywords:
electronic health recordslonelinessnatural language processing

More Related Videos

Artificial Intelligence-Based System for Detecting Attention Levels in Students
06:37

Artificial Intelligence-Based System for Detecting Attention Levels in Students

Published on: December 15, 2023

Related Experiment Videos

Last Updated: May 24, 2026

Methodology for Establishing a Community-Wide Life Laboratory for Capturing Unobtrusive and Continuous Remote Activity and Health Data
11:21

Methodology for Establishing a Community-Wide Life Laboratory for Capturing Unobtrusive and Continuous Remote Activity and Health Data

Published on: July 27, 2018

Artificial Intelligence-Based System for Detecting Attention Levels in Students
06:37

Artificial Intelligence-Based System for Detecting Attention Levels in Students

Published on: December 15, 2023

  • Explored large language model-based summarization of social and psychiatric history as an alternative input.

    • Main Results:

    • NLP models achieved modest accuracy (0.3-0.7) in classifying loneliness severity.
    • Models particularly struggled to identify severe loneliness, correlating with sparse and inconsistent documentation.
    • Summarization offered marginal accuracy improvements, mainly for mild loneliness predictions.
    • Manual review confirmed exceedingly rare explicit mentions of loneliness in clinical notes.

    Conclusions:

    • Current NLP model performance for loneliness classification is primarily limited by sparse and inconsistent EHR documentation.
    • The modeling approaches themselves are not the primary constraint.
    • Improving documentation practices is essential for advancing computational phenotyping of loneliness.