A New Automated Approach for Early Lung Cancer Detection with Improved Diagnostic Performance – A Preliminary Analysis
Keywords:Medical image Processing, Low-Dose Computed Tomography, Lung Cancer Detection, Computer-Aided Detection, Anisotropic Diffusion, Radiomic Image Features, Curvelet Transform and Support vector machine
Lung cancer is becoming the major cause of cancer-related deaths in human worldwide and Saudi Arabia is not an exception. Therein the identification of potentially malignant lung nodules is essential for the diagnosis and clinical management of lung cancer. Unfortunately, in clinical practice, however, interpretation of Computed Tomography (CT) images is challenging for radiologists due to the large number of cases. It is therefore extremely important task to develop computer aided diagnosis (CAD) systems that can aid and enhance the radiologist to potentially reduce false positive (FP) findings. Even though numerous methodologies are proposed for CAD system in the literatures, the one proposed in this work will definitely stand out in improving the sensitivity and specificity for the detection of small nodules particularly in low dose CT images. This work attempts to employ the powerful tool, radiomics quantitative imaging features within curvelet domain to detect and characterize lung nodules with improved sensitivity and specificity. Subsequently, support vector machine (SVM) is used to learn the 2D stochastic and 3D anatomic features of curvelet coefficients and classifies suspected regions either as malignant or benign based on geometric, texture and intensity. A preliminary analysis of the proposed methodology is presented and compared against the metrics, sensitivity, specificity and accuracy on publicly available LIDC database to serve as a benchmark for future research efforts.
How to Cite
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Re use and mixing of content policy- We follow Creative Commons Licence Policy. We follow CC BY. Please refer below for all details
This license lets others distribute, remix, adapt, and build upon our work, even commercially, as long as they credit us for the original creation.
- The journal allows readers to read, download, copy, distribute, print, search, or link to the full texts of its articles and allow readers to use them for any other lawful purpose.
- The journal allows the author(s) to hold the copyright without restrictions.
- The journal allows the author(s) to retain publishing rights without restrictions