Lung Cancer Detection Based On CT-Scan Images With Detection Features Using Gray Level Co-Occurrence Matrix (GLCM) and Support Vector Machine (SVM) Methods

Author(s):  
Qurina Firdaus ◽  
Riyanto Sigit ◽  
Tri Harsono ◽  
Anwar Anwar
2018 ◽  
Vol 5 (1) ◽  
pp. 24-30
Author(s):  
Fatema Tuj Johora ◽  
Mehdi Hassan Jony ◽  
Md Shakhawat Hossain ◽  
Humayun Kabir Rana

Lung cancer is one of the most dangerous diseases and prediction of it, is the most challenging problem nowadays. Most of the cancer cells are overlapped with each other. It is hard to detect the cells but also essential to identify the presence of cancer cells in the early stage. Early detection of lung cancer may reduce the death rate. In this study, we used the Grey Level Co-occurrence Matrix (GLCM) to extract the feature of cancer affected lung image and then Support Vector Machine (SVM) has been used to detect normal and abnormal lung cells after implementing the features. Our experimental evaluation using MATLAB demonstrates the efficient performance of the proposed system and in the result. GUB JOURNAL OF SCIENCE AND ENGINEERING, Vol 5(1), Dec 2018 P 24-30


Image classification is one of the major issues of image pre-processing approach. To resolve this issue a large number of classification approaches has been developed. In this work, a novel SVM-FA (support vector machine optimized with firefly approach) classifier is developed for detecting the lung cancer on the basis of the CT images. Lung cancer is considered one of the most critical and vital. Thus the early analysis of such kind of disease is required. For this purpose, the study implements the image pre-processing (filtration and segmentation) techniques to the input CT scan images. Then the SVM classifier, optimized with firefly approach is applied to the pre-processed data. The target of the work is to enhance the accuracy in the final prediction or output. For evaluating the proficiency level of the proposed SVM-FA approach, a comparison analysis is also performed in this work. The comparison is done among proposed work, traditional work and SVM classifier. On the basis of the obtained facts and figures, the proposed work is found to be effective and efficient in terms of the accuracy (96%) and specificity (83.333%) respectively


2019 ◽  
Vol 36 (4) ◽  
pp. 339-344 ◽  
Author(s):  
Tulasi Sajja ◽  
Retz Devarapalli ◽  
Hemantha Kalluri

Sign in / Sign up

Export Citation Format

Share Document