Year: 2026 | Month: May | Volume: 13 | Issue: 5 | Pages: 440-447
DOI: https://doi.org/10.52403/ijrr.20260540
Advanced CNN Based Lung Cancer Stage Prediction with Visualization Using Heat Maps and Report Generation
Pamula Kamakshi1, Perni Guna Anuhya1, Mangalagiri Krishna Gayathri2, Badiga Chaitanya3, Pullagora Likhitha4, Yeddula Dhanush Reddy5
1,2,3,4,5Department of Information Technology, Dhanekula Institute of Engineering & Technology, JNTUK, Vijayawada, Andhra Pradesh, India.
Corresponding Author: Perni Guna Anuhya
ABSTRACT
Lung cancer is one of the major causes of cancer-related deaths worldwide, and early detection of the tumor stage is very important for providing effective treatment by this, rate of survival will be increased. This work presents a deep learning based approach for automatic lung cancer stage prediction and severity analysis using medical images such as CT and X -ray images. The model combines two advanced convolutional neural networks architectures, ResNet50 and DenseNet121, in an ensemble framework so that both structural patterns and fine texture details from lung images can be effectively captured. By combining these two algorithms then the ensemble model has the capacity to perform each stage of analysis with greater accuracy than an individual deep learning framework. The system classifies the given image into different stages of lung cancer and also produces heat-map visualization to highlight the areas that are effected which helps the doctors to identify easily. In addition, a simple web based interface allows doctors to upload medical images and obtain stage prediction and severity report generation. In general, this work shows the potential of deep learning methods to help clinicians analyse medical images and make faster decisions in lung cancer detection.
Keywords: Lung Cancer Stage Prediction, Deep Learning, Ensemble CNN, ResNet50, DenseNet121, Medical Image Analysis, CT Scan Classification, Severity Analysis, Heat Map Visualization, Automated Report Generation.
[PDF Full Text]