Logistic Segmentation and Multistage Categorization based Predictive Modeling of Lung Cancer

The incidence of lung cancer is rising, and the fatal rate is beyond the measure. The treatment of cancer can be Successful if the condition is spotted at the earlier stage and found to be tiny and has not spread. Smokers and ex-smokers who are in the symptom-free stage are encouraged to get screened for lung cancer. X-rays are used to make cross-sectional images of the chest so as to do diagnosis of lung cancer. Symptoms to diagnose the disease are often non-existent and error prone in the early stages, making prediction is extremely challenging. Several computer-aided diagnostic methodologies and systems have been proposed, developed, and produced to detect the size and location and these methods leverage computer technology in various ways. The challenging characteristics of the lung cancer include high variability nodule classification and low decision making. In this paper machine learning based Multistage Categorization of Standard Segmentation (ML-MCSS) model has been proposed to assist clinicians deal with ambiguous pulmonary nodules found on the spot or through screening. Furthermore, the disease prediction accuracy of newly presented Logistic Segmentation Algorithms (LSA) is compared to that of the benchmark and data collected manually. The experimental result provides the step-by-step analyses of each method and their collective shortcomings were highlighted.