Identification of ketotic cattle based on milk spectroscopic characteristics: a chemometric modelling approach built upon the principles of Aquaphotomics

The early diagnosis of metabolic disorders in intensive farms, such as ketosis, allows a timely health treatment, resulting in improved animal welfare and reduced productivity and economic losses for farmers. This approach represents an opportunity for more sustainable animal production practices. For these reasons, we developed a chemometric processing model based on GNU Octave capable of identifying milk sample categories based on the chemo-physical characteristics of their near-infrared signature. Such a characterisation relies on the innovative approach of aquaphotomics, which investigates the interaction between light and water molecules in any biological system. Aquaphotomics leverages the properties of water as a biomarker and amplifier to identify changes in the system indicative of a perturbation, such as the onset of a metabolic disorder. The realised package allows spectral data elaborations, including preprocessing methods like the calculation of spectrum derivatives, noise removal and scatter correction techniques, as well as chemometric analysis like Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), and Partial Least Squares Regression (PLSR). In this way, the code enables the identification of the most significant water bands implicated in the origin of the disease, with the potential to represent an innovative approach for monitoring the health status of animals.