DeepWeb: A Web Application for Advanced Lung Cancer Identification Using Novel CNN Model

Lung Cancer is one of the deadliest forms of cancer that affects people which increases mortality rates worldwide. Identifying anomalies in medical photographs at the very early stage is crucial for both quantitative image study and patient care. So, in our research, we present a fully automated web application to predict lung cancer early on CT scan images using a robust deep learning model called Convolutional Neural Network (CNN). We also performed transfer learning models i.e. ResNet50 and VGG16 on the same dataset. Still, these models could not achieve higher accuracy than CNN as it showed an excellent performance achieving 99% accuracy and 97% of F1score value. The proposed model is established in three phases: image augmentation is done initially as preprocessing steps on the Kaggle IQ-OTH lung cancer dataset, followed by lung cancer classification using the CNN model, and finally, this model is used to create a web application. Our web application developed using Streamlit, can successfully predict normal, benign, and malignant classes of lung cancer showing the confidence score of each image inserted by the user. This prompt user experience with our web application based on the robust Convolutional Neural Network will assist medical professionals in enhancing the automated early screening and diagnosis of lung cancer as well as raising the patient survival rate effectively.