Sensitivity and specificity of infrared thermography in detection of subclinical mastitis in dairy cows

Abstract

The objectives of this experiment were to determine interrelationships among mastitis indicators and evaluate the subclinical mastitis detection ability of infrared thermography (IRT) in comparison with the California Mastitis Test (CMT). Somatic cell count (SCC), CMT, and udder skin surface temperature (USST) data were compiled from 62 Brown Swiss dairy cows (days in milk=117+/-51, milk yield=14.7+/-5.2 kg; mean +/- SD). The CORR, REG, and NLIN procedures of Statistical Analysis System (SAS Institute Inc., Cary, NC) were employed to attain interrelationships among mastitis indicators. The diagnostic merit of IRT as an indirect measure of subclinical mastitis was compared with CMT using the receiver operating characteristics curves. The udder skin surface temperature was positively correlated with the CMT score (r=0.86) and SCC (r=0.73). There was an exponential increase in SCC (SCC, x10(3) cells/mL=22.35 x e(1.31 x CMT score); R(2)=0.98) and a linear increase in USST (USST, degrees C=33.45+1.08 x CMT score; R(2)=0.75) as the CMT score increased. As SCC increased, USST increased logarithmically [USST, degrees C=28.72+0.49 x ln(SCC, x10(3) cells/mL); R(2)=0.72]. The USST for healthy quarters (SCC <or=400,000 cells/mL; n=94) was different from that for subclinical mastitic quarters (SCC >400,000 cells/mL; n=135) (mean +/- SE; 33.45+/-0.09 vs. 35.80+/-0.08 degrees C). The sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, positive predictive value, and negative predictive value were 95.6, 93.6, 14.97, 0.05, 95.0, and 93.6, respectively, for IRT and 88.9, 98.9, 83.56, 0.11, 99.2, and 86.1, respectively, for CMT. The area under the receiver operating characteristics curve for IRT and CMT was not different. In conclusion, as a noninvasive and quick tool, IRT can be employed for screening subclinical mastitis via measuring USST, with a high predictive diagnostic ability similar to CMT when microbiological culturing is unavailable. However, the reliability of IRT among cows with different characteristics and those living under various environmental conditions remains to be determined.

Invited review: Sensors to support health management on dairy farms

“…Thermal cameras have been used for the detection of mastitis; however, these devices were hand-held (Berry et al, 2003;Colak et al, 2008;Hovinen et al, 2008;Polat et al, 2010). Whereas these papers have described an association between mastitis and udder skin temperature, no validated algorithms have been described.…”

Mastitis diagnosis in ten Galician dairy herds (NW Spain) with automatic milking systems

“…This could mean that either lowest levels of infection are difficult to detect by the sensors installed in the AMS, or that there may not have been a real infection and that the error was in our interpretation of the CMT test. CMT interpretation can be subjective (Polat et al, 2010), and this might result in false positives and negatives (Viguier et al, 2009) when not being executed correctly (Lam et al, 2009). Some authors argue that quarters with low CMT-scores in foremilk probably do not have clinical mastitis (Rasmussen, 2001) or they could be subclinical cases.…”

Implementation in breeding programmes

“…The latter include automatic image analysis (including infrared tomography) of animals for the quantification of (changes in) body tissue reserves (Coffey et al, 2003) or udder health (Polat et al, 2010) as well as sensor technology (Ipema et al, 2008). Moreover, associations between milk quality and energy balance have been estimated (Stoop et al, 2009).…”