Lung Cancer Segmentation and Detection using Cuckoo Search Evolutionary Algorithms in Early Stage
Gaurav Thakur, Vikas Wasson · 2021 International Conference on Smart Generation Computing, Communication and Networking (SMART GENCON) · 2021
In all around the world the main reason of death among people is the lung cancer. If the detection of lung cancer is done at the initial stage it can increase the probabilities of survival in humans. If this disease is detected on time, the complete 5-year rate of survival of patients injured with lung cancer from 14% to 49%. The detection approach of lung cancer that uses image processing is utilized in this research for classification in the presence of lung cancer as a CT scan. The designed model used some steps like pre-processing, image segmentation, and feature extraction along with classification. For segmentation K-means along with Cuckoo Search Algorithm (CSA) is used. Feature extraction technique is utilized for extraction purpose of features from lung Region of Interest (ROI) data by using Maximally Stable Extrenal Regions (MSER) and these features are used to train Convolutional Neural Network (CNN), which is used for testing later. At the end to see the type of lung cancer diseases are present by using classification technique of neural network named as Convolutional Neural Network. The work has been executed in MATLAB software and performance has been assessed in terms of True Positive Rate (TPR), True Negative Rate (TNR), False Positive Rate (FPR), False Negative Rate (FNR), Classification Accuracy, Error, and Execution Time.