Enhanced Detection of Surface Plasmon Resonance Microscope via Multiangle Illumination and Pretrained Deep Neural Network

Abstract

Surface plasmon resonance microscopy (SPRM) has proved to be a powerful tool for label‐free imaging, detection, and mass/size measurement of single isolated virus, cells, and nanoparticles. Herein, SPRM that can be extended to detect the dense nanoparticles, such as counting the number density of nanoparticles, is demonstrated. To improve the counting accuracy, multiangle illumination is adopted in a conventional SPRM, then a deep learning method based on pretrained neural network approach is proposed to deal with the large number of SPRM images. Numerical simulations show that the new approach can significantly enhance the accuracy in detecting the number densities of nanoparticles. The ability to precisely count the number of particles is important for the monitor and detection of environmental pollution because that the number density of the particulate matter is strongly related to the air quality and atmospheric visibility.