College of Automation, Hangzhou Dianzi University, Hangzhou, Zhejiang Province 310018, People's Republic of China.
School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai 200030, People's Republic of China.
Anal Chem. 2021 May 25;93(20):7399-7404. doi: 10.1021/acs.analchem.1c01493. Epub 2021 May 11.
The unique capability of surface plasmon resonance microscopy (SPRM) in single nanoparticle analysis has found use in various chemical and biological applications. While SPRM offers exceptional sensitivity, the statistical analysis of numerous nanoparticles has been extremely laborious and time-consuming. Herein, we presented an image processing software package for nanoparticle analysis in SPRM, which is empowered by a deep learning algorithm. This package enabled fully automated nanoparticle identification, digital counting, three-dimensional tracking of particle locations, and quantification of dwell time and Brownian motion properties. With a built-in image filtering process to improve the contrast, robust identification and analysis have been achieved from SPRM images of low refractive index nanoparticles. This software tool would largely promote the translation of SPRM technology into the digital sensing platform for high throughput sample screening.
表面等离子体共振显微镜（SPRM）在单个纳米颗粒分析中的独特能力在各种化学和生物应用中得到了应用。虽然 SPRM 具有出色的灵敏度，但对大量纳米颗粒的统计分析非常繁琐和耗时。在此，我们提出了一种用于 SPRM 中纳米颗粒分析的图像处理软件包，该软件包由深度学习算法提供支持。该软件包能够自动识别纳米颗粒、数字计数、三维跟踪粒子位置，并对停留时间和布朗运动性质进行定量分析。通过内置的图像过滤过程来提高对比度，从低折射率纳米颗粒的 SPRM 图像中实现了稳健的识别和分析。该软件工具将极大地促进 SPRM 技术转化为高通量样品筛选的数字传感平台。
