Image Analysis Techniques

All microscopy groups use image analysis techniques to some extent, as we need methods to process an image from raw image or video to whatever we decide to quantify and study. We use ImageJ/Fiji for standard tasks, DDM for dynamical information from time-series and develop our own code for more specialised tasks.

Differential Dynamic Microscopy (DDM)

DDM is a technique best summarised as doing Dynamic Light Scattering (DLS) on a microscope [1]. DDM takes in a video sample of a colloidal suspension (e.g. milk or coffee) and, by a few assumptions, calculate the system intermediate scattering function. This is the same as we calculate with the established DLS method, which is an established commercial technique to perform particle sizing and microrheology.

DDM can give us the dynamical information with a much more flexible setup than before, since microscopes are easier to combine with other techniques as optical tweezers [2] and velocimetry [3]. DLS requires a very dilute sample, while DDM can handle multiple scattering events [4] and even larger colloids than in DLS.

A prerequisite for DDM is a large memory allocation, which can be pointed out as the reason DDM was not invented earlier. These days, laptop computers with sufficient memory is able to perform the calculations in minutes.

Check out this summary that compares DDM with DLS, and what it is able to do!

References:

[1] Cerbino, Roberto, and Veronique Trappe. 2008. ‘Differential Dynamic Microscopy: Probing Wave Vector Dependent Dynamics with a Microscope’. Physical Review Letters 100 (18): 188102. https://doi.org/10.1103/PhysRevLett.100.188102.

[2] Peddireddy, Karthik R., Ryan Clairmont, Philip Neill, Ryan McGorty, and Rae M. Robertson-Anderson. 2022. ‘Optical-Tweezers-Integrating-Differential-Dynamic-Microscopy Maps the Spatiotemporal Propagation of Nonlinear Strains in Polymer Blends and Composites’. Nature Communications 13 (1): 5180. https://doi.org/10.1038/s41467-022-32876-y.

[3] Dienerowitz, Maria, Michael Lee, Graham Gibson, and Miles Padgett. 2013. ‘Measuring Nanoparticle Flow with the Image Structure Function’. Lab on a Chip 13 (12): 2359–63. https://doi.org/10.1039/C3LC00028A.

[4] Nixon-Luke, Reece, Jochen Arlt, Wilson C. K. Poon, Gary Bryant, and Vincent A. Martinez. 2022. ‘Probing the Dynamics of Turbid Colloidal Suspensions Using Differential Dynamic Microscopy’. Soft Matter 18 (9): 1858–67. https://doi.org/10.1039/D1SM01598B.

Video analysis

We develop customized video analysis codes when needed, which is used extensively in our projects with evaporating droplets and when analyzing the flow fields in our two-phase flow experiments. We use Fiji/ImageJ for more standardized tasks, but find ourselves developing our own video analysis procedures, customized to our particular experiments.

Image courtesy: I. Beechey-Newman, N. Kizilova, A.A. Hennig, E.G. Flekkøy, & E. Eiser, Confined colloidal droplets dry to form circular mazes, Proc. Natl. Acad. Sci. U.S.A. 122 (32) e2508363122, https://doi.org/10.1073/pnas.2508363122 (2025).

Chord length distributions

Chord length distributions are related to the structure factor of a system, and serve as an alternative to pore-size distributions and porosities. We have used chord length distributions to study the coherence lengths in spinodal decomposition, and are currently working on using these distributions to better understand the finger thickness and formation of our colloidal drying patterns.

Page updated

Google Sites

Report abuse