Early Detection of Mastitis by using Infrared Thermography in Holstein-Friesian Dairy Cows via Classification and Regression Tree (CART) Analysis

Abstract

Subclinical mastitis is an important udder disease that negatively affects both the animal health and reduces profitability in dairy farms. The increasing performance of thermal cameras over time and their usability in different areas increase their use in livestocks. Infrared thermography (IRT) technology is a noninvasive method that can estimate the surface temperature of objects. The objective of this study was to evaluate early detection of mastitis in Holstein-Friesian dairy cattle by using both udder surface temperatures (Tmax) from images obtained with the help of a FLIR One Pro thermal camera and some parameters such as Lab (CIE L*, a*, b*), HSB (Hue, Saturation, Brightness), RGB (Red, Green, Blue) by processing thermal images with the help of ImageJ program via classification and regression tree (CART) analysis. According to California Mastitis Test CMT by using CART analysis in this study, 64.9% of cows with udder surface temperature lower than 38.85 were healthy, and 73.3% of cows higher than 38.85 were determined as unhealthy. As for SCC, 77.6% of cows with udder surface temperature lower than 38.65 were healthy and 58.6% of cows with higher than 38.65 were determined as unhealthy. The areas under ROC (AUC) were found to be statistically significant in the diagnosis of subclinical mastitis. (P<0.01) The sensitivity and specificity of the CART algorithm for CMT and SCC diagnostic tests were 85.42%, 81.48% and 90.20%, 80.39%, respectively. There was no significant difference between SCC and CMT tests in the area under the ROC curve (P>0.05). As a result, IRT technology can be used as a useful diagnostic tool in the early detection of mastitis.