Hybrid Approaches for Identification of Lung Cancer Using Image Processing

Lung Cancer is one of the significant types of cancer in humans. In order to prevent this, the disease should be recognized at early stages and with accuracy which can be helpful for continuing procedures like diagnosis, therapy, and other treatment. This paper provides a study that produces an ensemble approach based on the combined algorithms of deep learning like convolutional neural network and algorithms of machine learning such as decision tree classifier, random forest, logistic regression, and K-nearest neighbors to identify the cancerous cells. The model is built to automatically extract the important characteristic features of the cells, to recognize the patterns that indicate cancer in the lungs. Our method involves preprocessing the visual data to increase the quality of the picture and train the model on these images. The proposed architecture is then optimized through various parameters to gain the highest levels of accuracy. The accuracy we got from the proposed method is 94.19%.