A Self-Adaptive Frequency Domain Network for Continuous Intraoperative Hypotension Prediction

2025-09-28

AI Summary

Intraoperative hypotension (IOH) is a critical risk factor for postoperative delirium and increased mortality, necessitating highly robust early prediction methods. Existing AI models struggle to jointly model time-frequency features, capture both short- and long-range temporal dependencies, and suppress noise inherent in physiological signals. To address these limitations, we propose SAFDNet—a Self-adaptive Frequency-domain Network—that pioneers the integration of adaptive spectral analysis with an interactive attention mechanism for noise-robust, multi-scale temporal dependency modeling. SAFDNet jointly leverages Fourier transform, adaptive thresholding for denoising, and frequency-domain feature-guided attention. Evaluated on multiple real-world clinical datasets, it achieves an AUROC of 97.3%, significantly outperforming state-of-the-art methods. The framework demonstrates superior accuracy, strong noise robustness, and practical feasibility for clinical deployment.

Technology Category

• Natural Language Processing: Safety and Robustness
• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Intraoperative hypotension (IOH) is strongly associated with postoperative complications, including postoperative delirium and increased mortality, making its early prediction crucial in perioperative care. While several artificial intelligence-based models have been developed to provide IOH warnings, existing methods face limitations in incorporating both time and frequency domain information, capturing short- and long-term dependencies, and handling noise sensitivity in biosignal data. To address these challenges, we propose a novel Self-Adaptive Frequency Domain Network (SAFDNet). Specifically, SAFDNet integrates an adaptive spectral block, which leverages Fourier analysis to extract frequency-domain features and employs self-adaptive thresholding to mitigate noise. Additionally, an interactive attention block is introduced to capture both long-term and short-term dependencies in the data. Extensive internal and external validations on two large-scale real-world datasets demonstrate that SAFDNet achieves up to 97.3% AUROC in IOH early warning, outperforming state-of-the-art models. Furthermore, SAFDNet exhibits robust predictive performance and low sensitivity to noise, making it well-suited for practical clinical applications.

Problem

• Predicting intraoperative hypotension to prevent postoperative complications
• Overcoming limitations in time-frequency analysis of biosignal data
• Developing noise-robust AI models for clinical warning systems

Innovation

• Adaptive spectral block with Fourier analysis
• Self-adaptive thresholding for noise mitigation
• Interactive attention captures long-short dependencies