Author Email: email@example.com
Cancer is the most prevalent high-flown disease in all countries. In all cancer types, lung cancer is the most mortal disease. Millions of people were die every year because of lung cancer. Early recognition of disease plays a protuberant role in cancer patients. Accurate prophecy of this disease swells up the survival rate. But accurate detection of lung cancer is very critical with the existing systems and also time consuming. To conquer this difficulty the hybrid method is proposed. Accessibility of present technology has proved the way to explore the genes and its alliances in a variety of ailments like lung cancer. In this paper, a hybrid approach is proposed where genetic optimization algorithm is used for detection of cancer in CT images along with SVM classification combined with novel feature selection technique. This method assists the doctors to discern the lung nodules perfectly at early stages
Cancer is the most dangerous disease concerning with abnormal growth of cells. These cells grow up uncontrollably and form as malignant tumors. These tumors spread and affect the organs within the body. Out of all the cancer diseases lung Cancer is utmost perilous diseases that cause a great number of fatalities in both men and women and also it has been recognized as a major health problem in worldwide. According to ACS (American cancer Society), in the year 2018 approximately 14% are lung cancer out of all cancer types . In 2014, lung cancer causes 1.6 million deaths annually and estimates that by 2030 it will attain just about 10 million deaths per year . Early diagnosis of this disease can avert the spread of cancer, fetches better treatment and increases the endurance rate. However, the computed tomography (CT) imaging modality is used to screen the lung cancer. CT scan forms several slices (detailed images) of lung . Moreover, to diagnose lung cancer still manual methods are used in clinical practices which are not as much as effective for radiologists and pathologists. Therefore, an effective cancer detection system based on image processing is of great interest for evaluating larger dimensional medical data related to disease. Image processing technology is a dynamic diagnostic tool for medical purposes and has been developed very much to investigate the fine details. So to detect lung cancer the image processing technology and efficient optimization techniques are essential. In this paper a hybrid strategy is proposed to identify lung lesion effectively and exact. The proposed method employs genetic algorithm as a segmentation method that identifies genes (lung nodule) precisely, support vector machine for classification to classify whether the detected part is malignant or not and LBP feature extraction is used to the extract the features of lung cancer. In this work also compared the results of proposed system with DWT technique
In this paper, a novel strategy is proposed to recognize the cancer extant in CT images of Lung. This strategy utilizes Genetic algorithm and obtained 90% accuracy. Thus the proposed framework performs fine and is vigorous against anatomical varieties of the lungs. This work demonstrated a productive and exact outcome for dividing the lung tumor in CT therapeutic images and it will help radio-significance as second assessment for the better analysis of lung malignant growth.
527 total views, 1 views today