Differential Dynamic Microscopy can be applied to Differential Interference Contrast images despite shadowing effects

TL;DR

This paper demonstrates that Differential Dynamic Microscopy (DDM) can be effectively applied to Differential Interference Contrast (DIC) microscopy images despite shadowing effects, validating its use for analyzing isotropic motion in such images.

Contribution

The study provides a mathematical framework showing that shadowing effects in DIC do not affect DDM analysis of isotropic motion, supporting its broader application.

Findings

DIC shadowing causes deformation in DDM matrices but does not impact isotropic motion analysis.

Mathematical description of DIC shadowing effects on DDM was developed.

Parameter fitting remains invariant for directed advection and diffusion in DIC images.

Abstract

During In Vitro Fertilisation (IVF), combining time-lapse contrast microscopy with suitable image processing techniques could facilitate non-invasive classification of oocyte health, driving oocyte selection to maximise success rates. One image processing technique, Differential Dynamic Microscopy (DDM), has been used in a variety of microscopy settings including bright field, dark field and fluorescence. However, in some applications, effects stemming from the choice of microscopy may invalidate underlying assumptions of DDM. Here, we study the DDM analysis of differential interference contrast (DIC) microscopy movies. DIC exhibits a characteristic shadowing effect that gives the illusion of a 3D appearance, which we show causes deformation of the output DDM matrix. We present a mathematical description of this deformation, and conclude that, when studying isotropic motion, no account of the DIC shadow needs to be considered during DDM analysis. We validate our conclusions with simulated particle data, and with DIC images of colloidal dispersions. Although this conclusion does not generally extend to anisotropic motion, we demonstrate that for directed advection and diffusion behaviour, parameter fitting invariance still holds. These results validate the current practice of applying DDM to DIC, and are a foundation for further exploration of DDM in other phase-contrast image datasets.