Efficient Detection of Lung Nodule using Stacked U-Net++ and Multi Layer Perceptron

Lung cancer may develop in the different regions of lung with numerous reasons. Lung nodule is the one of the parts in lung, which is abnormal tissue growth or white spot in lung region. It may range from less than 3cm or larger than 3cm and also it is crucial to note that either the lung nodule is benign (non-cancerous) which means harmless tissue o nodule is malignant (cancerous) a harmful tissue with cancer cell. The harmful causes of lung cancer can prevent by early detection. There are several steps should be followed to detect the lung nodule using CT scan image. The process begins with preprocessing for removing irrelevant noise in the CT image, this is followed by segmentation of nodule to find the white spot in lung using Stacked U-Net++.Feature extraction will be used to identify exact region and conclude with classification, which can be done by using multi- layer perceptron to classify the nodule as benign or malignant. The aim of the proposed model is to provide high precision, recall, F1-measure, sensitivity and reduce the false positives and run time for detection.