Statistical detection of nanoparticles in cells by darkfield microscopy

Abstract

In the fields of nanomedicine, biophotonics and radiation therapy, nanoparticle (NP) detection in cell models often represents a fundamental step for many in vivo studies. One common question is whether NPs have or have not interacted with cells. In this context, we propose an imaging based technique to detect the presence of NPs in eukaryotic cells. Darkfield images of cell cultures at low magnification (10×) are acquired in different spectral ranges and recombined so as to enhance the contrast due to the presence of NPs. Image analysis is applied to extract cell-based parameters (i.e. mean intensity), which are further analyzed by statistical tests (Student's t-test, permutation test) in order to obtain a robust detection method. By means of a statistical sample size analysis, the sensitivity of the whole methodology is quantified in terms of the minimum cell number that is needed to identify the presence of NPs. The method is presented in the case of HeLa cells incubated with gold nanorods labeled with anti-CA125 antibodies, which exploits the overexpression of CA125 in ovarian cancers. Control cases are considered as well, including PEG-coated NPs and HeLa cells without NPs.

Keywords: Darkfield microscopy; Gold nanoparticles; Image analysis; Statistical test.