(184w) Federated Hierarchical Tensor Networks: A Collaborative Learning Quantum AI-Driven Framework for Healthcare

Healthcare industries frequently handle sensitive and proprietary data, and due to strict privacy regulations, they are often reluctant to share data directly. In today's context, Federated Learning (FL) stands out as a crucial remedy, facilitating the rapid advancement of distributed machine learning while effectively managing critical concerns regarding data privacy and governance. The fusion of federated learning and quantum computing represents a groundbreaking interdisciplinary approach with immense potential for revolutionizing various industries ranging from healthcare to finance. In this work, we proposed a federated learning framework based on quantum tensor networks, which leverages the principles of many-body quantum physics. There are currently no known classical tensor networks implemented in federated settings. Furthermore, we investigated the effectiveness and feasibility of the proposed framework by conducting a differential privacy analysis to ensure the security of sensitive data across healthcare institutions. Experiments on popular medical image datasets show that the federated quantum tensor network model achieved a mean receiver-operator characteristic area-under-the-curve (ROC-AUC) between 0.91-0.98. Experimental results demonstrate that the quantum federated global model, consisting of highly entangled tensor network structures, showed better generalization and robustness and achieved higher testing accuracy, surpassing the performance of locally trained clients under unbalanced data distributions among healthcare institutions.
Advancing medical research through collaborative healthcare data sharing presents challenges stemming from data heterogeneity (in terms of formats, standards, and structures) and stringent privacy regulations. The concept of federated learning (FL) holds great promise for advancing medical research by enabling collaborative analysis of decentralized data and addressing challenges related to data heterogeneity, communication, security, technical complexity, and legal considerations. Initially, FL was designed for various domains, including edge and mobile device use cases [1], but it has recently gained momentum in the realm of healthcare applications. It allows for model training on edge nodes, enabling real-time insights and decision-making without the need for extensive data transfers to a central server.
In the past few years, quantum computing, especially distributed quantum computing, including quantum machine learning (QML), has made remarkable progress. Its extraordinary ability to leverage the combined power of distributed quantum resources surpasses the constraints of individual quantum
nodes [2]. In parallel, Tensor networks (TNs) are considered promising candidates for quantum machine learning (QML) architectures [3].
Considering that tensor networks can serve as representations for both neural networks and quantum circuits [4], it becomes a natural objective to investigate the intersection of these two fields in federated settings. The main motivation is to unlock the potential benefits of quantum tensor networks and federated learning in the healthcare sector. This includes creating improved diagnostic tools, enhancements in medical image analysis, and ensuring meaningful results for rare diseases when smaller healthcare institutions lack sufficient data to train an accurate predictive model.
In this paper, we proposed a quantum tensor networks-based federated learning framework, or FedQTN, for collaborative learning between multiple healthcare institutions, enabling QTNs to train on non-independent and identically distributed medical images. To validate our approach and its adaptability to heterogeneous data, we demonstrated the learning capability of FedQTNs on popular medical image classification datasets, including chest radiographs from the RSNA and NIH datasets, MRI brain scans from the Alzheimer's ADNI dataset, and CT-scans from the CT-kidney dataset.
Keywords: quantum machine learning, federated learning, distributed optimization, healthcare, medical imaging, classification, privacy-preserving
References
1. McMahan, B., Moore, E., Ramage, D., Hampson, S. & Arcas, B. Communication-efficient learning of deep networks from decentralized data. Artificial Intelligence And Statistics. pp. 1273-1282 (2017)
2. Cuomo, D., Caleffi, M., & Cacciapuoti, A. S. (2020). Towards a distributed quantum computing ecosystem. IET Quantum Communication, 1(1), 3-8.
3. Rieser, H. M., Köster, F., & Raulf, A. P. (2023). Tensor networks for quantum machine learning. Proceedings of the Royal Society A, 479(2275), 20230218.
4. Grant, E., Benedetti, M., Cao, S., Hallam, A., Lockhart, J., Stojevic, V., & Severini, S. (2018). Hierarchical quantum classifiers. npj Quantum Information, 4(1), 65.