PulmoCare: Lung Cancer Detection

Abstract

Lung cancer remains the world’s deadliest cancer, accounting for the highest number of cancer-related deaths globally due to late-stage detection that often results in a five-year survival rate of less than 20%.[1] This study introduces PulmoCare, an CNN-based screening system designed to bridge the gap between advanced technology and clinical practice by automating the identification of pulmonary nodules in Low-Dose Computed Tomography (LDCT) scans. The methodology integrates optimized Deep Convolutional Neural Networks (DCNN), such as VGG19, with a Genetic Algorithm (GA) for intelligent feature selection, a process that minimizes computational redundancy and enhances classification precision by addressing the difficulty of distinguishing nodules from vascular structures.[2] Experimental results indicate that the proposed framework achieves high detection accuracies, ranging from 92% to 96.25%, while providing diagnostic results within seconds to assist medical practitioners in resource-limited settings.[3] Although barriers regarding model interpretability and the \"black box\" nature of deep learning persist, this multimodal approach represents a transformative step toward personalized lung cancer care, offering a reliable and scalable solution to improve early diagnosis and patient survival rates

Introduction

PulmoCare is a deep learning–based system designed for early detection of lung cancer from CT scans. Lung cancer is often diagnosed late due to subtle early symptoms, making early detection critical for treatment success. Traditional diagnostic methods like X-rays, CT scans, biopsies, and radiologist analysis are effective but time-consuming and require specialized expertise. PulmoCare uses Convolutional Neural Networks (CNNs) to automatically analyze preprocessed lung CT images, classify them as cancerous or non-cancerous, and generate detailed diagnostic reports through a web interface.

The system was trained and tested on datasets such as LIDC-IDRI and LUNA16, achieving 85–92% accuracy, 93% precision, and 91% recall, with a low 4% false positive rate. PulmoCare significantly reduces diagnostic time, human error, and workload for clinicians while improving access to lung cancer screening, especially in resource-limited settings.

Challenges include limited training data, detection of very small nodules, and the need for clinical validation. Future enhancements could involve transfer learning, ensemble models, integration with hospital databases, cloud and mobile platforms, and expansion to detect multiple lung conditions for more comprehensive and real-time medical support.

Conclusion

The implementation of AI systems like PulmoCare and large multimodal models represents a transformative advancement in lung cancer care, enabling significantly faster and more accurate diagnostic results than traditional manual interpretation. Research indicates that Deep Convolutional Neural Networks (DCNN), particularly when optimized with techniques like genetic algorithms, consistently achieve peak accuracies of up to 96.25% by automatically extracting complex patterns from medical imaging. While persistent barriers regarding model interpretability, \"black box\" algorithms, and data scarcity remain, the transition from traditional machine learning to advanced hybrid deep learning architectures offers a reliable pathway toward personalized pulmonary oncology. Ultimately, these technologies empower healthcare professionals by reducing diagnostic workloads and providing a critical \"second reader\" to improve early-stage detection rates and patient survival outcomes.

References

[1] J. Zhong, Y. Wang, D. Zhu, and Z. Wang, “A Narrative Review on Large AI Models in Lung Cancer Screening, Diagnosis, and Treatment Planning,” PMC NIH, 2025 [2] R. Javed, T. Abbas, A. H. Khan, A. Daud, A. Bukhari, and R. Alharbey, “Deep learning for lungs cancer detection: a review,” Artificial Intelligence Review, vol. 57, no. 197, July 2024 [3] A. Elnakib, H. M. Amer, and F. E. Z. Abou-Chadi, “Early Lung Cancer Detection Using Deep Learning Optimization,” International Journal of Online and Biomedical Engineering (iJOE), vol. 16, no. 6, pp. 82–94, 2020 [4] G. Cai, Y. Cai, Z. Zhang, Y. Cao, L. Wu, D. Ergu, Z. Liao, and Y. Zhao, “Medical Artificial Intelligence for Early Detection of Lung Cancer: A Survey,” Technical Survey, 2024 [5] D. Ardila, A. P. Kiraly, S. Bharadwaj, B. Choi, J. J. Reicher, L. Peng, D. Tse, M. Etemadi, W. Ye, and G. Corrado, “End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography,” Nature Medicine, vol. 25, no. 6, pp. 954–961, 2019. [6] “Kaggle, Lung Cancer Image Dataset,” [Online]. Available: https://www.kaggle.com/datasets [7] National Cancer Institute, “Lung Cancer Screening (PDQ®)–Patient Version,” Cancer.gov, 2024. [8] (2002) The IEEE website. [Online]. Available: http://www.ieee.org/

Copyright

Copyright © 2026 Mr. Manav Waghela, Mr. Yashas Kulal, Mr. Varun Surve, Mr. Abdul Rehman, Mrs. Silviya Dmonte. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.