Classification of lung glandular cells for early detection of cancer using multiple color spaces and scale space catastrophe features
Publication Type:Journal Article
Source:Life Science Journal, Volume 10, Number 2, p.2971-2980 (2013)
One of the biggest challenges the world face today is the mortality due to Cancer. One in four of all diagnosed cancers involve the lung cancer. The lung cancer remains the most common cancer-related cause of death both in developed and developing countries due to inhaling cancer-causing substance such as tobacco. Screening test help doctors to find and detect cancer at early stages. Several methods such as MRI, chest-X rays, CT Scan, etc., are available for screening tests. For developing countries, the cost involved for early detection with the available methods is not affordable. This paper presents a novel low cost method to detect and classify lung glandular cells as benign or malignant (Cancer cells) using conventional pap stained sputum cytology images. The microscopic sputum images are preprocessed and analysis is restricted to cellular regions. For segmentation we use multiple color spaces and clustering algorithms: K-Means and Fuzzy C-Means. Scale Space based Catastrophe points are used as features and are classified using Support Vector Machine (SVM). We successfully classified the glandular cells as benign or malignant cells with an accuracy of 78.61%.
cited By (since 1996)0
Cite this Research Publication
Related Research Publications
- Automated lung cancer detection by the analysis of glandular cells in sputum cytology images using scale space features
- Segmentation of lung glandular cells using multiple color spaces
- Early descriptions and classification of cancer in the classical Ayurvedic texts
- Organ specific tumor markers: What's new?
- Canavalia vivosa lectin and its tissue binding pattern in squamous cell carcinoma of oral cavity