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

You might also read

Related Articles

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

Sort by
Same author

Toxicity assessment in preclinical histopathology via class-aware Mahalanobis distance for known and novel anomalies.

Scientific reports·2026
Same author

Trends and Challenges of the Modern Pathology Laboratory for Biopharmaceutical Research Excellence.

Toxicologic pathology·2024
Same author

Fine-tuning of Fgf8 morphogen gradient by heparan sulfate proteoglycans in the extracellular matrix.

Biophysical journal·2024
Same author

GPR180 is a new member of the Golgi-dynamics domain seven-transmembrane helix protein family.

Communications biology·2024
Same author

Monocyte-derived alveolar macrophages are key drivers of smoke-induced lung inflammation and tissue remodeling.

Frontiers in immunology·2024
Same author

Learning image representations for anomaly detection: Application to discovery of histological alterations in drug development.

Medical image analysis·2023

Related Experiment Video

Updated: Jan 2, 2026

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
09:47

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

1.6K

Deep learning enables pathologist-like scoring of NASH models.

Fabian Heinemann1, Gerald Birk2, Birgit Stierstorfer2

  • 1Drug Discovery Sciences, Boehringer Ingelheim Pharma GmbH & Co. KG, 88397, Biberach an der Riß, Germany. fabian.heinemann@boehringer-ingelheim.com.

Scientific Reports
|December 6, 2019
PubMed
Summary

Automated deep learning models using convolutional neural networks (CNNs) can now score liver disease features like ballooning, inflammation, steatosis, and fibrosis in non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis (NASH). These AI-driven scores show strong agreement with human pathologists.

More Related Videos

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

7.3K
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
05:33

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System

Published on: July 11, 2025

721

Related Experiment Videos

Last Updated: Jan 2, 2026

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches
09:47

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

1.6K
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
07:15

Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model

Published on: August 16, 2020

7.3K
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
05:33

Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System

Published on: July 11, 2025

721

Area of Science:

  • Hepatology
  • Computational Pathology
  • Artificial Intelligence in Medicine

Background:

  • Non-alcoholic fatty liver disease (NAFLD) and its progressive form, non-alcoholic steatohepatitis (NASH), represent significant global health challenges with limited therapeutic options.
  • Histopathological assessment of liver biopsies, evaluating features like ballooning, inflammation, steatosis, and fibrosis, is crucial for disease staging and treatment efficacy studies.
  • Current histological scoring is labor-intensive, subjective, and requires specialized expertise, creating a bottleneck in research and clinical practice.

Purpose of the Study:

  • To develop and validate an automated system for quantifying histopathological features in NAFLD/NASH using deep learning.
  • To demonstrate the utility of convolutional neural networks (CNNs) in analyzing whole slide images of liver sections for disease assessment.
  • To compare the performance of automated scoring with traditional pathologist-based scoring methods.

Main Methods:

  • Whole slide images of stained liver sections were analyzed using four specialized CNNs, each targeting a specific histopathological feature (ballooning, inflammation, steatosis, fibrosis).
  • The CNNs were trained on two different scales to provide high-resolution, continuous quantitative readouts for each feature.
  • Continuous scores were mapped to discrete, pathologist-like scores for direct comparison with human expert evaluations.

Main Results:

  • The automated deep learning system successfully generated continuous quantitative scores for ballooning, inflammation, steatosis, and fibrosis.
  • These continuous scores could be accurately mapped to established discrete scores, mimicking pathologist assessments.
  • The automated deep learning-based scores demonstrated good agreement with the scores assigned by a human pathologist, validating the AI model's performance.

Conclusions:

  • Convolutional neural networks offer a powerful tool for automating the quantitative assessment of histopathological features in NAFLD/NASH.
  • This deep learning approach provides a high-resolution, objective, and reproducible method for evaluating disease progression and treatment response.
  • Automated scoring has the potential to significantly enhance the efficiency and consistency of NAFLD/NASH research and diagnostics.