Search research articles

Search research articles

Related Experiment Videos

An optimized EfficientNetB0 framework with CLAHE-based preprocessing for accurate multi-class chest X-ray

Nagwa Yaseen Hegazy¹, Mohamed S Sawah^2,3

¹Information Systems Department, Faculty of Information Systems and Computer Science, October 6 University, Giza, 12585, Egypt. Nagwa.Yaseen.Cs@o6u.edu.eg.

Scientific Reports

|March 28, 2026

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

Ensemble neural network models for stability prediction and optimization of hydraulic structures considering uplift pressure and exit gradient.

Scientific reports·2026

Same author

E-RespiNet: An LLM-ELECTRA driven triple-stream CNN with feature fusion for asthma classification.

PloS one·2025

Same author

Improving air quality prediction using hybrid BPSO with BWAO for feature selection and hyperparameters optimization.

Scientific reports·2025

See all related 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

This study introduces an optimized EfficientNetB0 model for multi-label chest X-ray classification, improving diagnostic accuracy for co-occurring thoracic diseases. The framework offers a robust solution for computer-aided diagnosis, outperforming other models.

Area of Science:

Medical Imaging
Artificial Intelligence
Computer-Aided Diagnosis

Background:

Chest radiography is vital for diagnosing thoracic diseases.
Interpreting multiple co-occurring pathologies on X-rays presents a significant challenge.
Existing deep learning models often oversimplify this complexity.

Purpose of the Study:

To develop an optimized deep learning framework for accurate multi-label classification of chest X-rays.
To address the challenge of co-occurring pathologies in thoracic disease diagnosis.
To improve the performance of computer-aided diagnosis systems.

Main Methods:

An optimized EfficientNetB0 framework was developed using the NIH dataset.
The model incorporated CLAHE-based contrast enhancement and strategic class balancing.

Keywords:

CLAHE preprocessing Chest X-ray classification EfficientNetB0 Thoracic pathology detection Transfer learning

Related Experiment Videos

A comparative transfer learning strategy was employed to maintain multi-label complexity.

Main Results:

The proposed model achieved a macro-average AUC of 0.906 and recall of 0.824.
It outperformed DenseNet121 and MobileNetV2 in diagnostic performance.
Strong per-class discrimination was observed, notably for Pneumonia (AUC=0.950) and Cardiomegaly (AUC=0.946).

Conclusions:

The framework effectively balances learning capacity and generalization in multi-label chest X-ray classification.
It provides a robust and interpretable solution for detecting co-occurring thoracic pathologies.
This approach enhances computer-aided diagnosis for complex clinical scenarios.