Quantum-Assisted Neuro-Fusion Network For Cross-Modal Intelligence In Biomedical Imaging

Recent advances in biomedical imaging demand intelligent systems capable of integrating heterogeneous data sources such as MRI, CT, PET, ultrasound, and histopathological images to achieve accurate and robust clinical decision-making. This paper presents a Quantum-Assisted Neuro-Fusion Network (QANFN) for cross-modal intelligence in biomedical imaging, leveraging the complementary strengths of quantum-inspired computation and deep neural fusion architectures. The proposed framework employs modality-specific deep feature extractors followed by a quantum-assisted entanglement-aware fusion layer that models high-order inter-modal correlations using quantum superposition and probabilistic amplitude encoding principles. A neuro-fusion decoder with attention-guided learning further refines fused representations to enhance diagnostic sensitivity and specificity. Experimental evaluation on multi-modal biomedical imaging datasets demonstrates that QANFN achieves superior performance in terms of classification accuracy, Dice similarity coefficient, sensitivity, and robustness against noise and missing modalities when compared with conventional CNN-based and transformer-based fusion models. The results confirm that quantum-assisted neuro-fusion significantly improves cross-modal feature representation and diagnostic reliability, making the proposed framework a promising solution for next-generation intelligent biomedical imaging systems.