用于吞咽和语音检测的全聚氨酯纳米网摩擦电传感器

文献信息

中文摘要

背景： 连续且无创地监测吞咽和发声对医疗保健和人机交互很重要，但由于喉部区域会发生大变形和持续运动，这仍然具有挑战性。现有的可穿戴传感器通常依赖块状弹性体或多层薄膜，可能对自然喉部运动造成机械负担。在此，我们提出了一种全聚氨酯（PU）纳米网摩擦电传感器，它在喉结上方形成顺应性界面，用于吞咽和语音的多模态监测。多孔纳米网界面直接接触皮肤，无需中间凝胶或块状层，从而保持透气性和保形接触。与代表性的连续聚二甲基硅氧烷薄膜相比，纳米网器件的水蒸气透过率约高14倍，并且在模拟喉结压缩过程中所需的机械功仅为13%。热压辅助稳定处理保持了PU纳米网的高拉伸性，同时相对于初纺纳米网，将初始电极电阻降低了85%，并将30%应变下的电阻增加抑制了80%。该器件可检测频率为几赫兹的吞咽相关喉部运动，以及100-300 Hz范围内的语音相关振动，声学共振接近150 Hz。与紧凑型无线模块集成后，该平台捕获了对应于喉部抬高和下降的W形吞咽信号，并在颈部姿势变化时保留了可测量的发声信号，信噪比为9.0-21.4 dB。这些结果建立了一种透气且机械顺应性的纳米网喉部界面，用于多模态喉部传感。

英文摘要

Continuous and non-invasive monitoring of swallowing and voice production is important for healthcare and human-machine interaction, yet remains challenging because the laryngeal region undergoes large deformation and continuous motion. Existing wearable sensors often rely on bulk elastomers or multilayer films that can impose mechanical burden on natural throat motion. Here, we present an all-polyurethane (PU) nanomesh triboelectric sensor that forms a compliant interface above the laryngeal prominence for multimodal monitoring of swallowing and voice. The porous nanomesh interface directly contacts the skin without intermediate gel or bulk layers, preserving breathability and conformal contact. Compared with a representative continuous polydimethylsiloxane film, the nanomesh devices exhibited an approximately 14-fold higher water-vapor transmission rate and required 13% of the mechanical work during laryngeal-prominence-mimicking compression. Heat-press-assisted stabilization maintained the high stretchability of the PU nanomesh while reducing the initial electrode resistance by 85% and suppressing the resistance increase at 30% strain by 80% relative to as-spun nanomeshes. The device detects swallowing-associated laryngeal motion at frequencies of several hertz and voice-related vibrations in the 100-300 Hz range, with an acoustic resonance near 150 Hz. Integrated with a compact wireless module, the platform captured W-shaped swallowing signals corresponding to laryngeal elevation and depression and retained measurable phonation signals under changing neck postures, with signal-to-noise ratios of 9.0-21.4 dB. These results establish a breathable and mechanically compliant nanomesh-based throat interface for multimodal laryngeal sensing.