Lung Cancer Detection Using CT Images and CNN Algorithm
Sushama Garud, Sudhir Dhage
Abstract
Cancer treatment is possible if we can able to detect it at an early stage. Generally, Symptoms of cancer are seen in the human body in the last stage, but with the help of advanced technology in which computer-aided systems are used, we can detect it in an early stage. Currently, numerous machine learning techniques are used for such automated detection systems to detect lung cancer in early stages. For such automated detection, we used convolution neural network (CNN) and Computed Tomography (CT) images. CT images are used due to their property of having less noise disturbance compared to MRI, X-Ray. On such CT scans, median filtering is used to improve the image quality. Such prepossessed images are then passed through the Alex net (25 layers) CNN architecture. Different layers of this architecture perform the task of feature extraction and classification. In the feature extraction part different low-level and different high-level features are extracted. The classification layer is responsible to detect whether the image provided is having a malignant or benign type of tumor or a normal image.