Automated Lung Cancer Detection using Prewitt Gradient Kernel & SVM from CT-Lung Images

It is well acknowledged that lung cancer is one of the leading causes of mortality globally. The scientific community has worked diligently to develop methods for detecting lung cancer at an early stage. A practical method of assessing the state of the lungs on the inside is via computed tomography (CT) pictures of the chest. When applied to high- resolution photos, image analysis may detect subtle changes and distortions that a human eye would miss. For early lung cancer diagnosis, several systems based on image analysis have been presented. Domains have a low accuracy and a high false- positive rate, despite this. To extract the form of cancer cells and blood vessels, the suggested technique makes use of Prewitt edge detection. In addition, the proposed approach employs support vector machines (SVM) for grading the normality or abnormality of a cell on the spot. The IQ-OTHNCCD dataset, used to evaluate the proposed system, has 1097 pictures split evenly between benign, malignant, and normal conditions. When compared to previously built systems, this one maintained a high degree of accuracy.