Pulmonary Chest Nodule Detection through Adaptive Reinforcement Learning Model (ARLM)

Lung cancer has become one of the serious problems in recent days and impact many lives. Lung cancer needs to be diagnosed accurately and in the early stages to avoid life straightening problems. Various factors are collaborated with the detection of lung tumor. Lung tumor can be treated in the initial stages using CT images MRI images clinical data X-rays and physiological data measurement. Any changes happening in the lung is impacted to the other organs of the body and spread the infections. In the existing system Semi supervised model is developed in which deep transfer learning approach is implemented. The existing system focused on developing lung cancer classification system in which malignant pulmonary nodal and benign cancer symptoms are detected. Virus in the propose to system had up to reinforced learning algorithm is developed. The chest images are collected from lung CT images of pre-processor and feature extracted using case feature extraction technique. The adaptive reinforced learning model incorporates with auto aggressive neural network provides repeated cross validation processes that strengthen the decision on tumor presents. The results are evaluated using accuracy and sensitivity. The objective of the proposed approach to detect the lung cancer presence in the early stages. False negative issues occurring in the existing development of convolutional architectures need to be rectified here with adaptive learning approach hence the false negative rate is being reduced with repeated operations. The proposed ARLM model achieved 92% accuracy on static image analysis.