Integration of Deep Learning Based Image Analysis in Lung Cancer Pathology

The recognition, prognosis, and medical treatment planning of lung cancer heavily depend on segmenting whole slide histopathology image analysis. Our research focused on analyzing lung tissue Whole Slide Images (WSI) using deep learning to get the best accuracy and reduce errors. This is especially important for analyzing medical images, such as detecting or separating tumors from healthy tissue. In this paper, we do annotation, data cleaning, data augmentation, and splitting. This process creates a collection of about 200,000 separate pictures that are ready to be studied and manipulated. We applied deep learning models like Fully convolutional network (FCN), and Convolutional Neural Network (CNN). Machine learning models such as clustering for grouping normal patches and tumor patches, as well as Principal Component Analysis (PCA), are also used. We have achieved 85.05% accuracy for the CNN model which gives us the highest accuracy. Histopathology image analysis to detect lung cancer and extract the features is important in medical education and helps to control cancer in society.