Advanced imaging for microalgal biotechnology

Abstract

The efficient operation of large-scale microalgae cultivation facilities requires continuous awareness of the culture condition. Although many conventional methods can be implemented in the laboratory, this goal can only be accomplished with non-invasive techniques such as spectral imaging based on the measurement of light backscattering of the culture surface. Several imaging methods are available, but we argue that developments in spectral approach will be among the essential breakthroughs for future advanced industrial-scale cultivation of microalgae.

The spectral methods initially developed for long-range (satellite and airborne) remote sensing of large water bodies are now increasingly employed for close-range monitoring of phytoplankton in natural ecosystems and large-scale microalgal cultures in open ponds and in closed photobioreactors. Similarly to high-throughput phenotyping which is now central to the progress of plant sciences, accelerated breeding, and precision farming, spectral imaging is gaining attention in microalgal biotechnology. Its power stems from the automated, rapid, non-invasive collection of large datasets, and the current advances in Machine Learning (ML). Their benefits include affordability, high information payload, and simplicity.

This review briefly presents imaging methods currently used in microalgal research, then focuses on spectral imaging. The background and biophysical foundation of remote sensing of communities and artificial monocultures is presented. Then, we elaborate on the methods for extracting relevant information from spectral images for monitoring of biomass accumulation, culture health, and target metabolites. Special attention was given to novel, trendy applications of ML to processing images and spectral data for the inference of actionable insights into the culture condition.