Method of studying the mechanisms of three-dimensional self-assembly of colloidal particles with the use of light-sheet microscopy and machine learning
Anton I. Shvetsov1, Egor V. Yakovlev1, Ivan V. Simkin1, Anastasia A. Shirokova1, Aleksandra V. Kokhanovskaia1, Aksiniia A. Bondareva1, Polina A. Zabavina1; 1Bauman Moscow State Technical University, Moscow, Russia
Abstract
In this work, we present an approach to the visualization and analysis of three-dimensional self-assembly of colloidal particles in liquid with the use of light-sheet microscopy, followed by the detection of their scattering profiles using the trained YOLOv8 neural network. The reconstruction of colloids’ coordinates in 3D space was implemented with the developed method based on k-means clustering of the scattering profile’s geometric parameters and subsequent recovery of the power-law dependence between the Z-coordinate and the area of scattering profile using the conducted calibration experiments in which, in addition to the main video camera and laser plane located in the XY plane, we used video camera and laser plane (with a different wavelength) in the XZ plane. The tracking of the colloidal particles was carried out using Crockier-Grier algorithm. It has been demonstrated that the proposed post-processing method allows for the identification of characteristic features of scattering profiles of colloidal particles and the extraction of their spatial arrangement with high accuracy (mean average precision exceeds 0.93). This approach has the potential to monitor the self-assembly processes of colloidal particles in 3D with high special and temporal resolution and is important for the development of new materials and technologies, including 3D bio-printing and micro- and nanofabrication.
This work was supported by the Russian Science Foundation, Grant No. 22-72-10128.
Speaker
Anton I. Shvetsov
Bauman Moscow State Technical University
Russia
Discussion
Ask question
