Ensemble DL for Improved Diagnosis of Respiratory Diseases Using YOLO and CheXNet on Chest X-rays

Abstract

This work discusses an ensemble deep learning model that combines You Only Live Once (YOLO) and CheXNet to discover different respiratory diseases based on chest X-ray samples. YOLO will be best at detecting objects in real-time to precisely locate the presence of an abnormality, whereas CheXNet, a DenseNet model that was pretrained on ChestX-ray14, is the best at classifying various lung pathologies. The advantage of this will be to tap into the strengths of these two models, making the diagnosis more accurate and improving clinical decision-making. The performance of the ensemble model was assessed on a diverse subset of annotated chest X-rays covering various diseases like pleural effusion, nodules,and masses. The average performance of the model over the main pathologies was a mean average precision of 0.72, while sensitivity and specificity were 0.81 and 0.80, respectively, which is a measurable improvement over the individual models. The results show that the sensitivity, specificity, and general accuracy of the individual models significantly improved. This is not only an innovative method of streamlining the diagnosis process but also offers a credible tool to radiologists,particularly in a setting where an expert’s interpretation is not easily available. Our findings indicate the possibility of transforming the world of respiratory disease diagnostics with the help of artificial intelligence-based models and request additional research and confirmation in clinical practice.

Received: 4 November 2025 | Revised: 3 July 2026 | Accepted: 13 August 2026

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

Data Availability Statement

The data that support the findings of this study are openly available in the National Institutes of Health at https://nihcc.app. box.com/v/ChestXray-NIHCC?utm_source=chatgpt.com.

Author Contribution Statement

Mayada A. Ahmed: Conceptualization, Methodology, Software, Validation, Formal analysis, Data curation, Writing – original draft, Supervision, Project administration. Rashid A. Saeed: Conceptualization, Methodology, Validation, Formal analysis, Data curation, Writing – original draft, Writing – review &editing, Supervision, Project administration. Mamoon M. Saeed: Conceptualization, Methodology, Validation, Formal analysis, Data curation, Writing – original draft, Writing–review &editing, Supervision, Project administration. Ahmed M. Giha: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Visualization. Tamahi A. Alhaj: Conceptualization, Methodology, Software, Validation, Investigation, Resources, Data curation, Visualization. Ammar Y. Mohamed: Conceptualization, Methodology, Software, Validation, Investigation, Resources, Data curation, Visualization. Tayba M. Alfadeil: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Visualization.