Search research articles

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: Dec 30, 2025

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Published on: June 21, 2024

A deep convolutional neural network architecture for interstitial lung disease pattern classification.

Sheng Huang¹, Feifei Lee², Ran Miao¹

¹Department of Control Science and Engineering, University of Shanghai for Science and Technology, Shanghai, China.

Medical & Biological Engineering & Computing

|January 23, 2020

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

Bioinspired Ruthenium Arene Complex with Pseudo-Vacant Coordination Sites as Efficient Small-Molecular Antioxidant Enzyme Mimics for Preventing Vascular Restenosis.

Research (Washington, D.C.)·2026

Same author

From heart to brain: a Case Report of individualized antithrombotic management for left ventricular thrombus and stroke in a young patient with acute myocardial infarction.

Frontiers in pharmacology·2026

Same author

Causal Links Between the Plasma Lipidome and Cognitive Performance: Targets, Pathways, and Lifestyle Influences.

Molecular neurobiology·2026

Same author

Short-term performance of bleeding risk scores in anticoagulated older patients with acute pulmonary embolism.

Thrombosis research·2026

Same author

Integrative multi-omics analysis reveals the domestication mechanism of black rice.

National science review·2026

Same author

Ag-Ti<sub>3</sub>C<sub>2</sub>T<sub>x</sub> MXene mediated H<sub>2</sub>O<sub>2</sub> reduction for electrochemical biosensing of kanamycin in real samples based on split aptamer engineering strategy.

Talanta·2026

Same journal

Robot-assisted laser osteotomy system based on dose-effect model-guided feedforward control.

Medical & biological engineering & computing·2026

Same journal

A randomized multi-window 3D deep learning approach for intracranial hemorrhage detection on non-contrast head CT.

Medical & biological engineering & computing·2026

Same journal

A novel SE-ResNet architecture for continuous estimation of wrist and hand movements from HD-sEMG.

Medical & biological engineering & computing·2026

Same journal

Anti-aliasing-enhanced WaveUNet for clinically reliable 12-lead ECG reconstruction from limited 3-lead input.

Medical & biological engineering & computing·2026

Same journal

Deep multi-modal features based spatio-temporal video regression for non-invasive hemoglobin estimation.

Medical & biological engineering & computing·2026

Same journal

Reduced mechanical strength correlates with decreased elastin content in aortic intima-media tissue: association with dissection in human ascending aortas.

Medical & biological engineering & computing·2026

See all related articles

This summary is machine-generated.

A new deep convolutional neural network (CNN) architecture and a two-stage transfer learning (TSTL) method improve interstitial lung disease (ILD) pattern classification. This approach effectively addresses limited training data challenges in medical imaging analysis.

Area of Science:

Medical Imaging
Artificial Intelligence
Pulmonology

Background:

Interstitial lung disease (ILD) encompasses diverse lung tissue inflammations requiring early diagnosis for effective treatment.
Accurate ILD diagnosis is challenging due to overlapping clinical presentations.
Computer-aided diagnosis systems and deep convolutional neural networks (CNNs) show promise in medical image analysis.

Purpose of the Study:

To develop a novel CNN architecture for classifying ILD patterns.
To propose a two-stage transfer learning (TSTL) method to overcome limited training data issues in ILD classification.
To leverage unlabeled lung CT data for improved diagnostic accuracy.

Main Methods:

Designed a new deep convolutional neural network (CNN) architecture for ILD pattern classification.

Keywords:

Convolutional neural networks (CNNs)Deep convolutional autoencoder Interstitial lung diseases (ILDs)Transfer learning

More Related Videos

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Published on: October 13, 2023

Related Experiment Videos

Last Updated: Dec 30, 2025

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Published on: June 21, 2024

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Published on: October 13, 2023

Developed a novel two-stage transfer learning (TSTL) strategy utilizing source and unlabeled lung CT data.

Employed unsupervised learning to optimize an objective function based on prediction confidence and mutual information for unlabeled data.

Main Results:

The proposed CNN architecture demonstrated high performance in ILD pattern classification.
The TSTL method significantly improved classification accuracy, outperforming existing approaches.
Knowledge learned from unlabeled lung CT data proved beneficial for the target domain.

Conclusions:

The novel CNN architecture and TSTL method offer a promising solution for accurate ILD pattern classification.
The TSTL strategy effectively mitigates the challenge of insufficient training data in medical image analysis.
This work highlights the potential of leveraging unlabeled data and transfer learning in diagnosing complex lung diseases.