SWIR turns moisture control in meat emulsions into a process issue

A study published on 5 September shows how short-wave infrared hyperspectral imaging (SWIR) could move moisture control in meat emulsions from laboratory testing towards rapid in-process verification. For cooked sausages, emulsified products and fine batters, moisture is not simply an analytical value: it affects yield, texture, water-holding capacity and batch consistency. The team prepared emulsions with different lean-to-fat ratios and mixing times from zero to 15 minutes. It recorded spectral images and compared them with moisture determination by oven drying. The researchers then compared partial least squares regression, random forest, Elastic Net and an ensemble model. According to the study, Elastic Net provided the strongest overall performance, while the combined model delivered stable, complementary predictions. The technical relevance is that the signal is not limited to free water. The important spectral regions were mainly associated with water absorption, but also carried fat- and protein-related information. Chemical images also indicated progressive homogenisation as mixing continued. The method may therefore capture not only the final formulation outcome, but also the physical state of dispersion within the batter. Plant-scale transfer has not yet been demonstrated. The work covered 84 samples under controlled conditions, and the authors call for validation using larger datasets and real industrial settings. The practical conclusion is therefore cautious but important: an in-line optical measurement will only add value when linked to recipe specifications, cutter time, batter temperature and analytical verification. Used that way, it could flag deviations before they become texture, yield or product-declaration problems.