细胞和组织的力学特性在决定生物功能中起着至关重要的作用，布里渊光谱技术作为一种弹性显微成像方法，可以无标记、非接触、高空间分辨率地表征活体细胞组织在生理和病理过程中的力学特征变化，为疾病的早期诊断提供了有力工具。为了进一步识别生物系统的快速粘弹性变化及其对外环境变化的动态反应，在低光学损伤下提高成像速度成为布里渊光谱显微成像面临的主要瓶颈；此外，为了更加灵敏地识别生物系统的微小力学性质差异，如何结合多种粘弹性对比机制进行高光谱分辨力的测量也是需要关注的问题。本论文以脉冲受激布里渊散射（Impulsive Stimulated Brillouin Scattering，ISBS）光谱测量和显微成像为研究对象，在增强粘弹性鉴别的灵敏度、提高测量速度和光谱分辨力、实现高空间分辨率的快速多信息显微成像等方面展开相关研究。 提出基于ISBS的声速、声衰减和弹光系数综合测量方法，拓展了ISBS方法的测量能力，提高了粘弹性鉴别的灵敏性。在建立完整光谱理论模型和粘弹性量化关系的基础上，一次毫秒量级的光谱测量即可得到包括弹性、粘性、弹光系数的丰富特征。与现有仅通过弹性进行表征的方法相比，这种多信息的表征方法能够实现更加灵敏的粘弹性鉴别。 提出矩阵束光谱分析方法。得益于其噪声鲁棒性高、对光谱失真免疫的特点，以低光学损伤实现了0.1 ms积分时间的快速测量，测量准确度与原有光谱分析方法相比提高了一个数量级。此外，针对多组分混合物的单点测量，通过预处理并引入噪声阈值提高了矩阵束方法的自适应能力，无需先验信息即可提高光谱分辨力和弱信号检测灵敏度。对于布里渊频移差仅为2.7 MHz的两种材料的混合物，检测灵敏度提高了一倍。 提出基于相位补偿和功率最优化设计的信噪比增强方法，在低光学损伤下实现了亚毫秒积分时间、十微米空间分辨率的多信息三维ISBS显微成像。在信噪比优化的基础上，结合矩阵束方法实现了相对标准差为0.26%的高精度测量。通过将探测光斑减小至声波长极限，使空间分辨率提高到横向十微米量级、轴向百微米量级，并结合多种粘弹性对比机制实现了多信息的快速三维显微成像。
The mechanical properties of cells and tissues play a critical role in determining biological functions. As a label-free, non-contact, and high spatial resolution imaging method, Brillouin spectroscopy has the capability to characterize the mechanical features of live cells and tissues during physiological and pathological processes, providing a powerful tool for early disease diagnosis. However, improving imaging speed under low optical damage and using rich mechanical contrast mechanisms to more sensitively identify rapid biomechanical changes and dynamic responses remain major challenges for Brillouin spectroscopy. This paper focuses on impulsive stimulated Brillouin scattering (ISBS) spectroscopy and imaging, aiming to enhance the sensitivity of mechanical characterization, improve the measurement speed and spectral resolution, and achieve multifunctional and fast imaging with high spatial resolution. A multifunctional method based on ISBS for sound speed, attenuation, and elasto-optic coefficient is proposed, which extends the measurement capability of the ISBS method and improves the sensitivity of viscoelastic characterization. By establishing a complete theoretical model of the spectrum and quantifying its relationship with viscoelasticity, rich properties including material elasticity, viscosity, and the elasto-optic coefficient can be obtained through a millisecond-level spectral measurement. Compared with the current method that only characterize elasticity, this multi-information method can achieve a more sensitive viscoelastic characterization. A matrix pencil spectral analysis method is proposed, which is robust to noise and immune to spectral distortion. With the matrix pencil method, the ISBS measurement achieves 0.1 ms integration time under low optical damage, and the measurement accuracy is improved by one order compared with the original spectral analysis method. In addition, for single-point measurement of multi-component mixtures, the adaptive capability of the matrix pencil method is improved, and the spectral resolution and sensitivity of weak signal detection can be enhanced without prior information. For mixtures of two materials with a frequency difference of only 2.7 MHz, the detection sensitivity is improved by a factor of two. A signal-to-noise ratio enhancement method based on phase compensation and power optimization design is proposed, and three-dimensional ISBS imaging with sub-millisecond integration time and ten-micrometer spatial resolution of rich mechanical information is achieved under low optical damage. Based on signal-to-noise ratio optimization, high-precision measurement with a relative standard deviation of 0.26% is achieved by combining the matrix pencil method. By reducing the probe spot size to the limit of the acoustic wavelength, the spatial resolution is improved to a ten-micrometer level in horizontal directions and a one-hundred-micrometer level in the axial direction, thus achieving three-dimensional imaging. Moreover, high-speed ISBS imaging with enhanced sensitivity is achieved by combining multiple mechanical contrast mechanisms.
