Non‐Destructive Quantification of Mycelial Biocomposite Growth Over Time

Mycelial biocomposites are sustainable alternatives to nonbiodegradable materials in building and packaging. Efficient manufacturing requires accurate, non-destructive quantification of growth over time, yet existing methods are often destructive or imprecise. This study develops and evaluates several non-destructive quantification methods for wood-flour biocomposites by using images to define mycelial density levels, low (no visible growth, removable surface hyphal coverage < 7.8%/cm²), medium (light growth, 7.8%-26.7%/cm²), and high (dense coverage, > 26.7%/cm²), and tracking changes in each level over time. The first method, the manual creation of masks for each growth level, provided rapid but coarse classification, estimating 64.1% high growth after 16 days with 3.5% user variability. An algorithmic masking approach improved detail detection, increasing estimated high-growth coverage to 81.7% but also variability to 9.3%. A fully automated deep-learning model proved fastest and most consistent, yielding 77.8% high-growth coverage and eliminating intra-user variability (0%). The deep-learning method was then applied to assess the effects of substrate supplements, revealing distinct growth patterns for each-differences not captured by traditional methods. These results demonstrate the effectiveness of automated quantification for mycelial biocomposites, enabling reproducible, high-resolution, non-destructive monitoring of growth and providing a foundation for more precise engineering and wider adoption of these materials.