Single-Nanoparticle Orientation Sensing by Deep Learning

Abstract

This paper describes a computational imaging platform to determine the orientation of anisotropic optical probes under differential interference contrast (DIC) microscopy. We established a deep-learning model based on data sets of DIC images collected from metal nanoparticle optical probes at different orientations. This model predicted the in-plane angle of gold nanorods with an error below 20°, the inherent limit of the DIC method. Using low-symmetry gold nanostars as optical probes, we demonstrated the detection of in-plane particle orientation in the full 0–360° range. We also showed that orientation predictions of the same particle were consistent even with variations in the imaging background. Finally, the deep-learning model was extended to enable simultaneous prediction of in-plane and out-of-plane rotation angles for a multibranched nanostar by concurrent analysis of DIC images measured at multiple wavelengths.

Nanoparticle Shape Determines Dynamics of Targeting Nanoconstructs on Cell Membranes

“…In situ monitoring of nanoconstruct-cell interactions using NPs as probes can deconvolute the effects of NP shape and protein corona in physiologically relevant conditions. However, because of the intrinsic low optical contrast compared to cells, ligand-functionalized organic NPs and DNA origamis cannot function as probes without conjugation to dyes or inorganic nanoparticles. − In contrast, gold nanoparticles (AuNPs) can be tracked via optical methods, because of their strong scattering properties, and also have advantages as probes because they are biocompatible, can be synthesized into various shapes, , and can be covalently functionalized with diverse ligands. − Anisotropic AuNPs such as gold nanostars (AuNS) show angle-dependent patterns in differential interference contrast (DIC) microscopy − with potential for 3D orientation tracking, which can provide information on ligand–receptor binding and endocytosis at the molecular level.…”

“…17−19 In contrast, gold nanoparticles (AuNPs) can be tracked via optical methods, because of their strong scattering properties, 20 and also have advantages as probes because they are biocompatible, can be synthesized into various shapes, 21,22 and can be covalently functionalized with diverse ligands. 23−25 Anisotropic AuNPs such as gold nanostars (AuNS) show angle-dependent patterns in differential interference contrast (DIC) microscopy 26−28 with potential for 3D orientation tracking, 29 which can provide information on ligand−receptor binding and endocytosis at the molecular level.…”

Optical Microscopy Systems for the Detection of Unlabeled Nanoparticles

“… 69 Another study established a computational imaging platform to determine the 3D orientation of anisotropic optical nanoprobes based on acquired DICM images. 73 The model predicted the orientation of nanorods with a high degree of accuracy and has the potential to enable a robust, rapid, and fully automated tracking of particle rotations during interaction with living cells.…”

Intelligent metaphotonics empowered by machine learning

“…It may become an extremely useful assisting tool for experimental measurements. The demonstrations in this field include characterization of orientation [204] or size [205] of metallic nanoparticles using measured spectral data. ML find its applications in a variety of microscopy and imaging techniques [206,207] as well as for tracking [208], localization [209] and analysis of single molecules [210].…”