Published Paper Details:

This project presents a novel hybrid framework integrating Multi- Constrained Joint Non-Negative Matrix Factorization (MC-JNMF) with Convolutional Neural Networks (CNNs) to study lung metastasis in soft tissue sarcomas using imaging and genomic data. By jointly analyzing high-dimensional data from medical images (such as CT or MRI scans) and corresponding genomic information, the proposed method seeks to identify critical latent features that contribute to the detection and understanding of lung metastasis. The MCJNMF frameworks integrate multiple constraints to better reflect the biological relationships between the two modalities, while CNNs automatically extract imaging features to improve prediction accuracy. This hybrid approach aims to improve clinical decisionmaking by providing more interpretable models and more accurate predictions regarding lung metastasis development in sarcoma patients. Preliminary results indicate that the integration of imaging and genomic data enhances the model's performance compared to traditional methods.