Two novel methods of testing microalgae cell wall integrity using image based segment anything model or optical density spectrum analysis

Abstract

Abstract The cost of disrupting microalgal cells to access valuable intracellular compounds is an industrial bottleneck. Screening for species with low cell integrity could reduce this cost, but existing suitable rapid methods are limited. This study investigated two novel approaches to estimate microalgal cell integrity of various species; An image-based analysis using Segment Anything Model (SAM) to quantify cell counts and size distributions from microscopy images, and a full-spectrum optical density analysis to track changes during cell disruption. Cell disruption was induced by sonication and quantified using an exponential decay rate constant ( k ) as the primary comparison metric, with D₉₅ (time to 95% disruption) included as an intuitive measure. SAM showed good agreement with the reference method (i.e., counting cell sized particles with a particle counter after varying sonication durations), though with a reduced sensitivity for smaller cells. The optical method captured species ranking but deviated in absolute k values, reflecting species-dependent optical effects. Comparing how easy cells rupture in terms of decay rate k (using reference method) for the species Tetraselmis suecica ( k = 0.097 min −1 ), Isochrysis galbana ( k = 0.255 min −1 ), Nannochloropsis oceanica (k = 0.002 min −1 ) and Chaetoceros gracilis ( k = 7.706 min −1 ) reveals Chaetoceros gracilis as easily disrupted and Nannochloropsis oceanica was almost not affected.

The authors' abstract, as published at the source. Journal of Applied Phycology, 2026 · DOI ↗