Intelligent Recommendation System for Early Detection and Care Support in Alzheimer’s Disease

This paper presented an attention-based multimodal fusion framework for early Alzheimer’s disease detection. The proposed approach integrates heterogeneous clinical inputs and assigns adaptive weights to each modality, allowing the model to focus on the most informative features for prediction. The framework is designed to improve diagnostic accuracy, robustness, and interpretability compared with simple single-modality or concatenation-based methods. [2], [4], [5] The main contribution of the work is a structured diagnosis-support pipeline that can help identify Alzheimer’s disease at an early stage. By combining multimodal clinical data with attention-based fusion, the system is aligned with current research trends in multimodal AI for Alzheimer’s diagnosis and prognosis. Future work will focus on incorporating additional modalities such as MRI, EEG, PET, speech, and biomarker data to improve diagnostic performance. Explainable AI methods can also be added to make the system more transparent for clinicians. In addition, future studies should evaluate the framework under missing-modality conditions, cross-site datasets, and longitudinal records to improve generalization, fairness, and progression analysis. [1], [4], [11]

[1] Z. Yu, A. Mulholland, T. Huang, and Q. Liu, “Multimodal AI for Alzheimer Disease Diagnosis: Systematic Review,” JMIR, 2026.  [2] Y. Tu, S. Lin, J. Qiao, Y. Zhuang, and P. Zhang, “Alzheimer’s disease diagnosis via multimodal feature fusion,” Computational Biology and Medicine, vol. 148, p. 105901, Sep. 2022, doi: 10.1016/j.compbiomed.2022.105901.  [3] F. Leony and C.-J. Lin, “Multimodal fusion architectures for Alzheimer’s disease diagnosis: An experimental study,” Journal of Biomedical Informatics, vol. 166, p. 104834, Jun. 2025.  [4] X.-A. Bi, R. Cai, Y. Wang, and Y. Liu, “Effective Diagnosis of Alzheimer’s Disease via Multimodal Fusion Analysis Framework,” Frontiers in Genetics, vol. 10, art. 976, Oct. 2019, doi: 10.3389/fgene.2019.00976.  [5] M. Zhang et al., “A feature-aware multimodal framework with auto-fusion for Alzheimer’s disease diagnosis,” Computational Biology and Medicine, vol. 178, p. 108740, Aug. 2024, doi: 10.1016/j.compbiomed.2024.108740.  [6] B. Lei et al., “Alzheimer’s disease diagnosis from multi-modal data via feature inductive learning and dual multilevel graph neural network,” Medical Image Analysis, vol. 97, p. 103213, Oct. 2024, doi: 10.1016/j.media.2024.103213.  [7] J. Wang et al., “Deep joint learning diagnosis of Alzheimer’s disease based on multimodal feature fusion,” BioData Mining, vol. 17, no. 1, p. 48, Nov. 2024, doi: 10.1186/s13040-024-00395-9.  [8] J. Qin et al., “3D Multimodal Fusion Network With Disease-Induced Joint Learning for Early Alzheimer’s Disease Diagnosis,” IEEE Transactions on Medical Imaging, vol. 43, no. 9, pp. 3161–3175, Sep. 2024, doi: 10.1109/TMI.2024.3386937.  [9] “Clinical classification metrics for imbalanced medical datasets,” article title.  [10] “Explainable AI for healthcare decision support,” article title.  [11] “Attention-based multimodal fusion with contrast for robust clinical prediction,” article title.