AI-Driven Livestock Biosensing for Prediction of Metabolic Diseases

We report the development of a highly sensitive 3D-printed sensor for the on-farm, early detection of subclinical hypocalcemia (SHC) in dairy cows. The printed 3D sensing structure incorporates periodic micropatterns of ion-to-electron polymer-based transducing layer that enhances sensitivity when analyzing milk samples. This novel sensor detects radiometric targets of calcium (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{Ca}^{2+}$</tex>) and phosphate (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{PO}_{4}{}^{2-}$</tex>) in milk, enabling the identification of SHC in under a minute. We apply regression models, including <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{k}$</tex> Nearest Neighbors (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$k$</tex>-NN) and Logistic Regression, to predict livestock health, evaluating performance through accuracy, area under the curve (AUC), and confusion matrices. Unlike traditional tests, this sensor provides dairy farmers with a tool to monitor the health of transition dairy cows.