Deep Learning Based Classification of Lung Cancer Images: A Comparative Performance Evaluation

The purpose of this study is to use deep learning techniques to categorize lung cancer histological images into benign and malignant groups. Both benign and malignant lung histopathology images are included in the dataset. For binary classification, two well-known convolutional neural network models have been used: VGGNet-12 and ResNet50. In order to evaluate the performance of the models, various metrics have been used. The metrics are accuracy, loss, F1 score, precision, recall, confusion matrix. Additionally, two models have been evaluated to see which one performs better in classifying histological images of lung cancer. VGGNet achieved an accuracy of 99%, whereas ResNet achieved an accuracy of 95%. The comparison's findings provide information on how well these deep learning methods play a major role in processing of medical images.