Related Experiment Video
Updated: Jul 20, 2025
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
14.2K
Identification and Classification of Lungs Focal Opacity Using CNN Segmentation and Optimal Feature Selection
Muhammad Ashar Javed1, Hannan Bin Liaqat2, Talha Meraj3
1Department of Information Technology, University of Gujrat, Gujrat, Pakistan.
Computational Intelligence and Neuroscience
|August 4, 2023
Summary
Early lung cancer detection is crucial for survival. This study introduces a predictive model using semantic segmentation and optimal features, achieving 97.8% accuracy in identifying lung nodules.
Area of Science:
- Medical Imaging
- Oncology
- Computer-Aided Diagnosis
Background:
- Lung cancer has a high mortality rate, with survival significantly improving upon early detection.
- Accurate identification of lung nodules is challenging due to visual similarities with surrounding tissues, leading to misclassification.
- Previous studies often used noisy features, compromising diagnostic accuracy.
Purpose of the Study:
- To develop an accurate predictive model for detecting and classifying lung nodules.
- To improve early lung cancer diagnosis through advanced image analysis techniques.
- To address limitations of previous methods by utilizing optimal features and robust classification.
Main Methods:
- Employed semantic segmentation to precisely identify lung nodules within the Lungs Image Database Consortium (LIDC) dataset.
- Extracted optimal features including Histogram Oriented Gradients (HOGs), Local Binary Patterns (LBPs), and geometric features post-segmentation.
- Utilized Support Vector Machines (SVM) as the primary classifier for nodule identification.
Main Results:
- The proposed model achieved a high accuracy of 97.8% in detecting and classifying lung nodules.
- Demonstrated excellent sensitivity of 100% and specificity of 93% in nodule identification.
- Reported a low false positive rate of 6.7%, indicating robust performance.
Conclusions:
- The developed predictive model significantly enhances the accuracy of lung nodule detection and classification.
- Support Vector Machines proved effective for nodule identification, outperforming other classifiers.
- This approach holds promise for improving early lung cancer diagnosis and patient outcomes.
