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 Concept Videos

Classification of Illness01:17

Classification of Illness

8.5K
The meaning of illness is individualized to each person who experiences an alteration in health. In contrast, disease is a medical term indicating a pathological change in the structure and function of the body or mind. It is a condition that has specific symptoms and boundaries.
An illness is a response to a disease in which the person's level of functioning is changed compared with a previous level. The general classification of illness includes acute and chronic.
Acute illness is severe...
8.5K

You might also read

Related Articles

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

Sort by
Same author

Simultaneously achieving high sensitivity and long-term chromatic stability in anthocyanin-loaded PVA/chitosan nanofibers for real-time food spoilage monitoring.

Food research international (Ottawa, Ont.)·2026
Same author

Endoscopic closure of a malignant entero-vaginal fistula using a cardiac septal occluder device.

Endoscopy international open·2026
Same author

KlimaNot-Effects of Climate Change on Emergency and Acute Care: Protocol for a Multicenter, Registry-Based Observational Cohort Study.

JMIR research protocols·2026
Same author

Spatial and multi-omic profiling reveals pericyte-derived CCL19 as a key prognostic factor in CNS lymphoma.

HemaSphere·2026
Same author

Prostate cancer and diabetes: A retrospective analysis of mortality trends in the United States (1999-2024).

Medicine·2026
Same author

Distinct systemic metabolic signatures in premenopausal women with lipedema revealed by composite indices.

Frontiers in endocrinology·2026

Related Experiment Video

Updated: Jan 8, 2026

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems
05:47

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems

Published on: June 13, 2025

1.2K

Optimized BERT-based NLP outperforms zero-shot methods for automated symptom detection in clinical practice.

Juan G Diaz Ochoa1,2, Natalie Layer3, Jonas Mahr1

  • 1QuiBiQ GmbH, Stuttgart, Germany.

Frontiers in Digital Health
|December 12, 2025
PubMed
Summary

Fine-tuned Natural Language Processing (NLP) models reliably extract clinical symptoms from German emergency department (ED) texts, outperforming zero-shot approaches. This enables structured analysis of patient symptom profiles, even with data protection constraints.

Keywords:
clinical NLPfine-tuninglarge language models (LLM)named entity recognition (NER)natural language processing (NLP)symptom extraction

More Related Videos

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

986
Machine Learning-Based Cough Tone Classification: Diagnostic Exploration of Chronic Obstructive Pulmonary Disease and Respiratory Tract Infections
06:22

Machine Learning-Based Cough Tone Classification: Diagnostic Exploration of Chronic Obstructive Pulmonary Disease and Respiratory Tract Infections

Published on: September 19, 2025

399

Related Experiment Videos

Last Updated: Jan 8, 2026

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems
05:47

Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems

Published on: June 13, 2025

1.2K
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

986
Machine Learning-Based Cough Tone Classification: Diagnostic Exploration of Chronic Obstructive Pulmonary Disease and Respiratory Tract Infections
06:22

Machine Learning-Based Cough Tone Classification: Diagnostic Exploration of Chronic Obstructive Pulmonary Disease and Respiratory Tract Infections

Published on: September 19, 2025

399

Area of Science:

  • Clinical Informatics
  • Natural Language Processing
  • Medical Data Analysis

Background:

  • Natural Language Processing (NLP) offers tools for extracting clinical information from unstructured medical narratives.
  • Extracting symptom information from German emergency department (ED) free texts is challenging due to documentation pressure, varied language, and strict data protection regulations.
  • German is a low-resource language for clinical NLP, necessitating specialized approaches.

Purpose of the Study:

  • To implement and compare NLP models for recognizing symptoms, anatomical terms, and negations in German ED anamnesis texts.
  • To evaluate the feasibility of on-premises, data-secure model deployment for clinical text analysis.
  • To establish a pipeline for systematic symptom extraction and transformation into structured data.

Main Methods:

  • Comparison of two zero-shot learning models (GLiNER, Mistral) and a fine-tuned BERT-based model (SCAI-BIO/BioGottBERT) for Named Entity Recognition (NER).
  • Utilized manual annotations of 150 narratives for model validation in an on-premises hospital environment.
  • Implemented postprocessing steps including confidence filtering, negation exclusion, symptom standardization, and integration with structured oncology data.

Main Results:

  • The fine-tuned SCAI-BIO/BioGottBERT model achieved an F1 score of 0.84 for symptom extraction, outperforming zero-shot models.
  • Demonstrated superior performance in detecting negations compared to zero-shot approaches.
  • Successfully created a validated pipeline for extracting and structuring affirmed symptoms from ED free text.

Conclusions:

  • Modern NLP methods, particularly fine-tuned models, can reliably extract clinical symptoms from German ED free text under strict data protection rules.
  • This approach provides a precise and practical solution for integrating unstructured clinical narratives into decision-making.
  • The methodology supports large-scale analysis of patient cohorts and can be extended to extract other clinical entities, forming a basis for subgroup analysis.