Antifreezing Ultrathin Bioionic Gel-Based Wearable System for Artificial Intelligence-Assisted Arrhythmia Diagnosis in Hypothermia

Abstract

Cardiovascular disease (CAD) is a major global public health issue, with mortality rates being significantly impacted by cold temperatures. Stable and reliable electrocardiogram (ECG) monitoring in cold environments is crucial for early detection and treatment of CAD. However, existing skin sensor struggle to balance freeze resistance, breathability, flexibility, conductivity and adhesion at cold temperatures. Here, we introduce a solvent cross-linking strategy and an in situ transfer method to prepare ultrathin bioionic gels, featuring a freezing point below −80 °C and a thickness of only 12.6 μm. The strong and abundant interactions between the ionic liquid solvent and the zwitterionic polymer effectively suppress low-temperature crystallization, forming a toughened and highly adhesive network structure. This network enables the in situ formation of an ultrathin morphology, which can be seamlessly transferred onto various substrates. Furthermore, the solvent-cross-linked network maintains a large interpolymer chain spacing, facilitating rapid ion transport pathways. Even at subzero temperatures, the gel maintains its multifunctionality, demonstrating tissue-like softness (34.6 kPa), high ionic conductivity (10.06 mS cm^–1), excellent stretchability (360%), high transparency, robust adhesive strength (175.3 kPa) and interfacial toughness (1146 J m^–2). Integrated into a flexible wearable device, the ultrathin gel ensures excellent skin conformity, user comfort, and high signal-to-noise ECG signal acquisition. Leveraging an artificial neural network, the system analyzes bradycardia ECG signals and achieves 96.88% accuracy in arrhythmia detection under cold conditions. This bioionic gel-based system presents a promising solution for early CAD diagnosis and prediction in extreme environments.