A Dual-Stage AI Healthcare System for Brain Cancer Risk Prediction and Brain Tumor Detection

Abstract

This paper proposes a hybrid Vision Transformer–Convolutional Neural Network (ViT-CNN) framework for the early detection of Interstitial Lung Disease (ILD) using chest CT scans and X-ray images. ILD is a group of lung disorders characterized by inflammation and fibrosis that can lead to respiratory failure if not diagnosed early. Conventional diagnosis relies on manual interpretation of medical images, which is time-consuming, requires expert radiologists, and is prone to inter-observer variability, often missing subtle abnormalities in the early stages. While CNNs effectively extract local image features, they have limited ability to capture long-range spatial relationships, whereas Vision Transformers (ViTs) excel at learning global contextual information. The proposed hybrid model combines the strengths of both architectures to improve diagnostic accuracy and support clinicians with faster and more reliable decision-making. The system uses publicly available CT and X-ray datasets containing both healthy and ILD cases. Images undergo preprocessing steps including resizing, normalization, noise reduction, lung region enhancement, and data augmentation to improve model robustness and reduce overfitting. The CNN branch extracts local texture features such as fibrosis and ground-glass opacities, while the ViT branch captures global structural relationships across the lungs. The extracted features are fused and processed through fully connected layers to classify the presence and severity of ILD. The model is implemented using TensorFlow or PyTorch with supporting tools such as OpenCV and scikit-learn, and is deployed through a Flask or FastAPI backend with a React-based web interface, enabling clinicians to upload medical images and receive real-time predictions along with explainable Grad-CAM heatmaps. Experimental results demonstrate that the hybrid ViT-CNN model outperforms standalone CNN and ViT models in terms of accuracy, precision, recall, F1-score, and ROC-AUC. It achieves higher sensitivity for early-stage ILD detection, reducing false negatives and enabling earlier treatment. Visualization techniques such as Grad-CAM and attention maps confirm that the model focuses on clinically relevant lung regions, improving interpretability and clinician trust. The system also shows stable training, strong generalization to unseen data, and fast inference times suitable for real-world clinical deployment. Overall, the proposed framework provides an accurate, scalable, and explainable AI-based decision support system for early ILD diagnosis, with the potential to enhance patient outcomes and reduce diagnostic variability.

Introduction

This study proposes an AI-based healthcare framework for early brain cancer risk prediction and brain tumor detection by integrating machine learning with deep learning in a unified diagnostic platform. Existing research often focuses separately on clinical data analysis or MRI image processing, lacks early-stage predictive support, and provides limited model interpretability. To address these gaps, the proposed system combines clinical risk assessment with MRI-based tumor detection and segmentation, enabling more accurate and explainable diagnosis.

The framework consists of modules for patient data acquisition, risk prediction, tumor detection, report generation, and secure system administration. Machine learning models analyze patient demographics, medical history, and lifestyle factors to classify brain cancer risk, while deep learning models (CNN/U-Net) detect, classify, and segment tumors from MRI scans. The system generates comprehensive diagnostic reports with visual and quantitative analyses to assist clinicians in treatment planning. It is designed to be scalable, secure, interoperable, and suitable for cloud deployment, with features such as role-based access control and encrypted data management.

The proposed architecture follows a modular, multi-layer design comprising client, backend, AI processing, and data management layers, ensuring efficient real-time processing and maintainability. Experimental evaluation using the BraTS 2020 dataset demonstrated that the risk prediction module accurately classified patients into low-, medium-, and high-risk categories, while the tumor segmentation module achieved 89% segmentation accuracy and a Dice score of approximately 0.88, outperforming conventional 3D U-Net approaches in accuracy while requiring lower computational resources. The results indicate that the proposed system provides an effective, interpretable, and computationally efficient solution for early brain cancer diagnosis and clinical decision support.

Conclusion

The proposed Dual-Stage AI Healthcare System for Brain Cancer Risk Prediction and Brain Tumor Detection represents a significant step toward improving early diagnosis and clinical decision support through intelligent automation. By integrating machine learning–based risk assessment with deep learning–based MRI tumor analysis, the system addresses major limitations of traditional diagnostic workflows such as delayed detection, manual interpretation, and fragmented data analysis. The unified framework enables automated processing of clinical and imaging data, providing accurate, consistent, and interpretable results for healthcare professionals. The study conducted in this phase focused on requirement analysis, architectural design, and system modelling. Detailed examination of user requirements, diagnostic workflows, and technology feasibility ensured that the proposed architecture remains scalable, secure, and modular. The incorporation of predictive analytics for risk evaluation and convolutional neural network–based segmentation for tumor detection establishes a strong foundation for an intelligent and context-aware healthcare platform capable of supporting real-time clinical environments. From a design perspective, the multi-layered architecture—comprising client interface, backend services, AI processing modules, and medical data storage—provides a structured framework that supports extensibility and future integration of advanced healthcare technologies. Careful modelling of system workflows, data handling mechanisms, and interaction patterns ensures that the transition from design to implementation can be achieved with minimal architectural modification. The system architecture also supports cloud deployment and multi-institutional scalability, making it suitable for large-scale medical applications

References

[1] A. N. Mishra and P. K. Gupta, “HomeServe: The On-Demand Home Services Platform,” International Journal of Computer Applications, vol. 183, no. 24, pp. 12–18, June 2021. [2] S. Agarwal, R. Ghosh, and M. Sharma, “AI-Powered Service Recommendation System Using K-Means and NLP,” in Proc. 2022 IEEE International Conference on Computational Intelligence and Communication Technology (CICT), Ghaziabad, India, 2022, pp. 242–248. [3] A. Banerjee and S. Verma, “Fraud Detection in Digital Platforms Using Machine Learning: A Case Study,” IEEE Access, vol. 9, pp. 122601–122613, Sept. 2021. [4] H. Chen, R. H. Chiang, and V. C. Storey, “Business Intelligence and Analytics: From Big Data to Big Impact,” MIS Quarterly, vol. 36, no. 4, pp. 1165–1188, Dec. 2012. [5] T. Mikolov, K. Chen, G. Corrado, and J. Dean, “Efficient Estimation of Word Representations in Vector Space,” in Proc. International Conference on Learning Representations (ICLR), Scottsdale, USA, 2013. [6] R. S. Sutton and A. G. Barto, Reinforcement Learning: An Introduction, 2nd ed. Cambridge, MA, USA: MIT Press, 2018. [7] J. Han, M. Kamber, and J. Pei, Data Mining: Concepts and Techniques, 3rd ed. San Francisco, CA, USA: Morgan Kaufmann, 2011. [8] A. Gupta and V. Kumar, “Smart Service Marketplace: AI-Driven Local Service Recommendation and Scheduling,” in Proc. IEEE International Conference on Smart Computing (SMARTCOMP), Hong Kong, 2020, pp. 145–152. [9] M. F. Akbar, S. H. Memon, and R. Ali, “An Intelligent Framework for Digital Service Provider Verification Using AI,” IEEE Transactions on Emerging Topics in Computing, vol. 10, no. 2, pp. 580–591, 2022. [10] P. Resnick and R. Zeckhauser, “Trust Among Strangers in Internet Transactions: Empirical Analysis of eBay’s Reputation System,” The Economics of the Internet and E-commerce, vol. 11, pp. 127–157, 2002.

Copyright

Copyright © 2026 Dr. K. C. Nalavade, Nisarg Ketankumar Pansora, Atharv Shyam Gawali, Amit Ashok Arote, Charulata Sanjay Sonar. 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.