Search research articles

Related Concept Videos

Ultrasound II: Endoscopic Ultrasound and FibroScan

Ultrasound II: Endoscopic Ultrasound and FibroScan

Endoscopic Ultrasound (EUS) and FibroScan are valuable diagnostic tools in gastroenterology and hepatology, each with specific applications and techniques.
Endoscopic Ultrasound (EUS):

You might also read

Related Articles

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

Sort by

Same author

The genome sequence of the solitary wasp, <i>Nysson trimaculatus</i> (Rossi, 1790) (Hymenoptera: Crabronidae).

Wellcome open research·2026

Same author

ColonCrafter: A Depth Estimation Model for Colonoscopy Videos Using Diffusion Priors.

Pacific Symposium on Biocomputing. Pacific Symposium on Biocomputing·2026

Same author

The genome sequence of the Square-spot Deerfly, <i>Chrysops viduatus</i> (Fabricius, 1794) (Diptera: Tabanidae).

Wellcome open research·2026

Same author

The genome sequence of the Dune Robberfly, <i>Philonicus albiceps</i> (Meigen, 1820).

Wellcome open research·2026

Same author

The genome sequence of a tephritid fruit fly, <i>Xyphosia miliaria</i> (Schrank, 1781) (Diptera: Tephritidae).

Wellcome open research·2026

Same author

The Endo-GeneScreen platform identifies drug-like probes that regulate endogenous protein levels within physiological contexts.

Nature communications·2025

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Jul 9, 2025

Author Spotlight: Establishing MASLD Cell Models for Investigating Disease Mechanisms and the Lipid-Lowering Effects of Koumiss

Author Spotlight: Establishing MASLD Cell Models for Investigating Disease Mechanisms and the Lipid-Lowering Effects of Koumiss

Published on: July 19, 2024

Improving nonalcoholic fatty liver disease classification performance with latent diffusion models.

Romain Hardy¹, Joe Klepich¹, Ryan Mitchell¹

¹School of Information, U.C. Berkeley, Berkeley, CA, USA.

Scientific Reports

|December 7, 2023

Summary

Synthetic images from diffusion models improve nonalcoholic fatty liver disease (NAFLD) classification, even with limited real data. This approach enhances diagnostic tools for medical professionals.

More Related Videos

Author Spotlight: A Non-Invasive Tool to Assess and Differentiate Fat Patterns in Liver Using 3D Dixon MRI

Author Spotlight: A Non-Invasive Tool to Assess and Differentiate Fat Patterns in Liver Using 3D Dixon MRI

Published on: October 20, 2023

The Dimethylnitrosamine Induced Liver Fibrosis Model in the Rat

The Dimethylnitrosamine Induced Liver Fibrosis Model in the Rat

Published on: June 17, 2016

Related Experiment Videos

Last Updated: Jul 9, 2025

Author Spotlight: Establishing MASLD Cell Models for Investigating Disease Mechanisms and the Lipid-Lowering Effects of Koumiss

Author Spotlight: Establishing MASLD Cell Models for Investigating Disease Mechanisms and the Lipid-Lowering Effects of Koumiss

Published on: July 19, 2024

Author Spotlight: A Non-Invasive Tool to Assess and Differentiate Fat Patterns in Liver Using 3D Dixon MRI

Author Spotlight: A Non-Invasive Tool to Assess and Differentiate Fat Patterns in Liver Using 3D Dixon MRI

Published on: October 20, 2023

The Dimethylnitrosamine Induced Liver Fibrosis Model in the Rat

The Dimethylnitrosamine Induced Liver Fibrosis Model in the Rat

Published on: June 17, 2016

Area of Science:

Medical Imaging
Artificial Intelligence in Healthcare
Machine Learning for Diagnostics

Background:

Deep learning offers potential for improved medical diagnostics but requires extensive annotated data.
The scarcity of annotated medical images hinders the application of machine learning models in healthcare.
Nonalcoholic fatty liver disease (NAFLD) diagnosis can benefit from advanced computational methods.

Purpose of the Study:

To investigate the efficacy of using synthetic images generated by diffusion models to augment real medical images for NAFLD classification.
To compare the quality of synthetic images produced by diffusion models versus Generative Adversarial Networks (GANs).
To assess the performance enhancement in NAFLD prediction using synthetic data augmentation in low-data scenarios.

Main Methods:

Generated synthetic medical images using diffusion models and compared them with GAN-generated images.
Evaluated synthetic image quality using Inception Score (IS) and Fréchet Inception Distance (FID).
Employed a partially frozen Convolutional Neural Network (CNN) backbone (EfficientNet v1) for NAFLD classification with synthetic data augmentation.

Main Results:

Diffusion-generated images demonstrated superior quality, achieving a higher maximum IS (1.90 vs. 1.67) and a lower minimum FID (69.45 vs. 100.05) compared to GANs.
The synthetic augmentation method, combined with a CNN, achieved a maximum ROC AUC of 0.904 for image-level NAFLD prediction.
Performance enhancement was observed even in low-data regime settings, highlighting the method's effectiveness.

Conclusions:

Diffusion models are effective in generating high-quality synthetic medical images suitable for data augmentation.
Synthetic image augmentation using diffusion models significantly improves NAFLD classification performance, particularly in data-scarce environments.
This research provides a viable strategy to overcome data limitations in medical AI, enhancing diagnostic capabilities.