The Use of Infrared Thermography for the Monitoring of Udder Teat Stress Caused by Milking Machines

SIMPLE SUMMARY: The aim of this study was to test the use of infrared thermography as a possible tool for detecting short-term stress, of cow udder teat, caused by milking procedures. Thermographic images were collected and evaluated to calculate the values of two indicators: the average and the maximum skin surface temperatures at the base, center, and tip of each teat. Obtained results confirmed a relationship between the two indicators (T(avg), T(max)) and the level of teat stress generally evaluated by visual observation of its color. Nevertheless, the low accuracy reached by the two indicators does not seem to justify the development of an ad hoc infrared device for the monitoring of cow udder teat stress. ABSTRACT: The aim of this study was to test infrared thermography (IRT) as a possible tool for scoring teat color changes after cluster removal; thus, indirectly, to classify the short-term stress of teats caused by milking machines. Thermographic images (n = 137) from three farms were collected and evaluated to calculate the average and maximum skin surface temperatures (SSTs) at the base, center, and tip of each teat (T(avg,B), T(avg,C), T(avg,T), T(max,B), T(max,C), and T(max,T)). Obtained results confirmed a significant relationship between the indicators T(avg), T(max) and the levels of teat color change (level one: pink-colored teat; level two: red-colored teat; level three: blue or purple-colored teat). Nevertheless, when a teat was considered to be stressed because its scoring fell in level 3 of the color-change scale used, sensitivity and specificity in the classification of the teat status ranged respectively between 45.6% and 54.3%, and 54.4% and 59.2%, for the indicators T(avg); and 56.5% and 60.9%, and 59.7% and 61.8%, for the indicators T(max). When a teat was considered stressed because its scoring fell between the levels 2 and 3 of the scale adopted, sensitivity and specificity were between 49.0% and 55.8%, and 58.3% and 61.8%, for the indicators T(avg); and 55.8% and 59.9%, and 60.6% and 61.4%, for the indicators T(max). As a consequence, the low values of sensitivity and specificity do not seem to justify the development of an ad hoc infrared device for the monitoring of udder teat stress. Nonetheless, this technology can be a viable solution for a preliminary evaluation of the mechanical stress of teats if a milking system would be equipped with an infrared sensor already in place for other purposes (e.g., the monitoring of udder health status).