Related Experiment Video
Updated: Jan 17, 2026
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
14.7K
Machine-learning model for differentiating round pneumonia and primary lung cancer using CT-based radiomic analysis
Hasan Genç1, Mustafa Yildirim2
1Department of Radiology, Elaziğ Fethi Sekin City Hospital, Elaziğ, Turkey.
Medicine
|September 17, 2025
Summary
Machine learning models accurately distinguish round pneumonia from lung cancer using computed tomography (CT) radiomic features. This noninvasive approach aids diagnosis and reduces unnecessary procedures.
Area of Science:
- Radiology
- Artificial Intelligence
- Oncology
Background:
- Round pneumonia mimics lung cancer on CT scans, complicating diagnosis.
- Accurate differentiation is crucial to prevent invasive procedures.
- Radiomics and machine learning offer potential for noninvasive diagnostic aids.
Purpose of the Study:
- To develop and validate machine learning models for differentiating round pneumonia from primary lung cancer.
- To leverage radiomic features from CT images for improved diagnostic accuracy.
Main Methods:
- Retrospective study of 48 patients (24 round pneumonia, 24 lung cancer).
- Extraction of 107 radiomic features from CT images.
- Feature selection using information gain, identifying 5 key features.
- Training and validation of 7 machine learning classifiers.
Main Results:
- Several models, including Naïve Bayes, achieved perfect classification (AUC=1.000).
- The Naïve Bayes model, after feature selection, showed high performance (AUC=1.000, accuracy=0.979, sensitivity=0.958, specificity=1.000).
Conclusions:
- CT-based radiomics and machine learning effectively differentiate round pneumonia from lung cancer.
- These models represent a promising noninvasive tool for radiological diagnosis.
- The approach can reduce diagnostic uncertainty and guide clinical decisions.
