An iterative method based FLC-SLM system design for forming multiple complex structures simultaneously in 3D volume with tissue

TL;DR

This paper introduces a novel iterative wavefront shaping system using a weighted-mutation genetic algorithm to rapidly form multiple complex 3D structures through turbid media, validated with tissue and diffuser experiments.

Contribution

It presents a cost-effective, robust wavefront shaping system with a new genetic algorithm for fast, high-contrast, multi-structure 3D formation in scattering media.

Findings

Achieved fast convergence with high contrast and resolution.

Successfully formed multiple complex structures simultaneously.

Validated robustness across various diffusers and tissue samples.

Abstract

Complex structure formation and fast focusing of light inside or through turbid media is a challenging task due to refractive index heterogeneity, random light scattering and speckle noise formation. Here, we have proposed a weighted-mutation assisted genetic algorithm (WMA-GA) with an R-squared metric based fitness function that advances the contrast, resolution, focuses light tightly and does fast convergence for both simple and complex structure formation through the scattering media. As a compatible system with the binary WMA-GA, we have presented a fast, cost-effective, and robust iterative wavefront shaping system design with an affordable ferroelectric liquid crystal (FLC) based binary-phase spatial light modulator (SLM). The proposed wavefront shaping system design has been used to construct multiple complex hetero-structures simultaneously in 3D volume by an optimized single phase-mask. The WMA-GA and the prototype system have been validated with 120, 220, 450, and 600 grit ground glass diffusers along with 323, 588, and 852 {\mu}m thick fresh chicken tissues including fluorescence in it. We have demonstrated the robustness of the proposed method to control the photon-in and photon-out from a localized fluorescent dye embedded in the tissue. The detailed results show that the proposed class of algorithm-backed integrated system converges fast with higher contrast and advances the resolution.