Search papers, labs, and topics across Lattice.
This study introduces a federated learning (FL) framework designed for privacy-preserving oral cancer screening on smartphones, addressing the challenges of data access in the medical domain. By leveraging decentralized datasets and secure networking through Tailscale, the framework enables collaborative model development among geographically distributed researchers without compromising patient privacy. The implementation of lightweight architectures, particularly the MobileNetV4-Conv-S model, resulted in a global model achieving an AUC of 0.929 and an accuracy of 87%, demonstrating the efficacy of FL in enhancing oral cancer detection capabilities.
Achieving an impressive AUC of 0.929, this federated learning framework revolutionizes oral cancer screening by enabling secure, decentralized collaboration without compromising patient privacy.
Data are the cornerstone of robust AI models. However, in the medical domain, access to reliable data is constrained by regulatory requirements and patient privacy, and clinical oral images are particularly difficult to obtain. Federated learning (FL) mitigates these constraints by enabling collaborative model development across decentralized datasets without centralizing or sharing patient data. This work presents a practical FL framework that supports geographically distributed collaboration among AI healthcare researchers and facilitates the development of robust models for oral cancer screening. Client devices were interconnected via Tailscale to provide secure networking and real-time communication. We implemented the FL workflow using the Flower framework for server-side aggregation, while client deployment and orchestration were configured manually; no enterprise FL platforms were used. To support a smartphone-based screening application, we evaluated lightweight, mobile-friendly architectures including MobileNetV2, MobileNetV3Large, and MobileNetV4-Conv-Small (MNv4-Conv-S). Across the global lightweight models aggregated using FedAvg, the MNv4-Conv-S based global model (GM-V4) achieved the best performance, reaching an AUC of 0.929 and an accuracy of 87%
