END-TO-END DEEP NEURAL NETWORK FOR CANCER TYPE CLASSIFICATION FROM LIQUID BIOPSIES

In this research, an End-to-End Deep Neural Network model was proposed for cancer type classification using liquid biopsy data. The system integrates multiple modules including data collection, preprocessing, feature extraction, deep neural network modeling, and classification within a unified framework. By directly processing raw genomic and molecular biomarker data, the proposed architecture eliminates the need for manual feature engineering and enables automatic learning of meaningful patterns associated with different cancer types. The experimental results demonstrate that the proposed model can accurately classify multiple cancer types with improved performance compared to traditional machine learning methods. The deep neural network successfully learns complex relationships within high-dimensional biomedical data and provides reliable predictions for cancer diagnosis. In addition, the use of regularization techniques and optimized training strategies improves the model’s generalization ability and reduces overfitting. The system offers a scalable and efficient solution for analyzing liquid biopsy data and supporting clinicians in early cancer detection and diagnosis.   Although the proposed approach shows promising results, there are several opportunities for further research and improvement. Future work may focus on integrating additional biological data sources such as gene expression profiles, proteomic data, and clinical patient information to enhance the model’s predictive capability. Advanced deep learning architectures such as transformer-based models and attention mechanisms can also be explored to further improve classification accuracy. In addition, implementing the system in real-time clinical decision support platforms could help healthcare professionals perform faster and more accurate cancer diagnosis. Expanding the model to detect early-stage cancers and rare cancer types using large-scale biomedical datasets will also be an important direction for future research.

[1]      Cherukuri, S. P., Kaur, A., Goyal, B., Kukunoor, H. R., Sahito, A. F., Sachdeva, P., et al., “Artificial Intelligence-Enhanced Liquid Biopsy and Radiomics in Early-Stage Lung Cancer Detection,” Cancers, vol. 17, no. 19, 2025.   [2]   Luo, X., Xie, S., Hong, F., Li, X., Wei, Y., Zhou, Y., et al., “From multi-omics to deep learning: Advances in cfDNA-based liquid biopsy for multi-cancer screening,” Biomarker Research, 2025. [3]    Li, L., Geng, Y., Chen, T., et al., “Deep learning model targeting cancer surrounding tissues for accurate cancer diagnosis,” Journal of Translational Medicine, vol. 23, 2025.   [4]  Shao, X., Shen, Y., Sun, P., et al., “Real-time deep learning for tumor segmentation and tool tracking in breast biopsy,” World Journal of Surgical Oncology, 2025. [5]  Eledkawy, A., Hamza, T., & El-Metwally, S., “Precision cancer classification using liquid biopsy and advanced machine learning techniques,” Scientific Reports, 2024.   [6]   Guerry, J., “Recent advances in liquid biopsy for precision oncology and clinical applications in lung cancer,” Future Oncology, 2025.   [7]  Mishra, A. K., Alrubayan, M., & Pradhan, P., “Automated grading and staging of ovarian cancer using deep learning,” arXiv, 2025. [8]  Zang, Z., Dorward, D. A., Quiohilag, K. E., et al., “Deep learning-based pathological subtyping of non-small cell lung cancer,” arXiv, 2025.   [9]   Sarfati, E., Bône, A., Rohé, M., et al., “Deep learning-based classification of hepatocellular carcinoma using CT scans,” arXiv, 2025.   [10]  Di Caro, G., Kirakosyan, V., Abanov, A. G., et al., “Quantum cognition machine learning for forecasting chromosomal instability,” arXiv, 2025.