BVFLMSP : Bayesian Vertical Federated Learning for Multimodal Survival with Privacy

2026-04-02

AI Summary

This work addresses the challenges of privacy constraints limiting centralized training of sensitive data and the lack of predictive uncertainty quantification in multimodal survival analysis. To overcome these issues, the authors propose a Bayesian longitudinal federated learning framework based on Split neural networks. In this framework, each client models its local modality using a Bayesian neural network, while a central server aggregates intermediate representations to predict survival risk. Differential privacy is incorporated at the client side to ensure secure data transmission. This approach represents the first integration of Bayesian neural networks with longitudinal federated learning for multimodal survival analysis, enabling effective modality fusion and reliable uncertainty estimation under strict privacy guarantees. Experimental results demonstrate that the method achieves up to a 0.02 improvement in C-index across various modality combinations, significantly outperforming baseline approaches such as MultiSurv.

Abstract

Multimodal time-to-event prediction often requires integrating sensitive data distributed across multiple parties, making centralized model training impractical due to privacy constraints. At the same time, most existing multimodal survival models produce single deterministic predictions without indicating how confident the model is in its estimates, which can limit their reliability in real-world decision making. To address these challenges, we propose BVFLMSP, a Bayesian Vertical Federated Learning (VFL) framework for multimodal time-to-event analysis based on a Split Neural Network architecture. In BVFLMSP, each client independently models a specific data modality using a Bayesian neural network, while a central server aggregates intermediate representations to perform survival risk prediction. To enhance privacy, we integrate differential privacy mechanisms by perturbing client side representations before transmission, providing formal privacy guarantees against information leakage during federated training. We first evaluate our Bayesian multimodal survival model against widely used single modality survival baselines and the centralized multimodal baseline MultiSurv. Across multimodal settings, the proposed method shows consistent improvements in discrimination performance, with up to 0.02 higher C-index compared to MultiSurv. We then compare federated and centralized learning under varying privacy budgets across different modality combinations, highlighting the tradeoff between predictive performance and privacy. Experimental results show that BVFLMSP effectively includes multimodal data, improves survival prediction over existing baselines, and remains robust under strict privacy constraints while providing uncertainty estimates.

Problem

• Multimodal Survival Analysis
• Privacy Preservation
• Time-to-Event Prediction
• Uncertainty Quantification
• Vertical Federated Learning

Innovation

• Bayesian neural networks
• Vertical Federated Learning
• Multimodal survival analysis
• Differential privacy
• Uncertainty quantification