Relation between infrared thermometry and thermal imaging for assessment of rodents’ body temperature: a way to improve animal welfare.

Abstract

Animals have been used for biomedical research since ancient times. Humane endpoints have been increasingly implemented during experimental protocols to prevent or alleviate animals’ suffering. Body temperature is an indicator of stress or anxiety. Infrared thermometer (IRT) and infrared thermal imaging camera (TIC) are commercially available for temperature measurement. This study aimed to determine mice and rats’ body temperature using IRT and TIC, and to establish a mathematical correlation between both approaches, to understand if it is possible to infer TIC readings from IRT readings in animals’ back and tail. Thirty Wistar rats (15 females and 15 males) and thirty FVB/n mice (15 females and 15 males) were used. The body temperature of each animal was measured for ten consecutive days using IRT (UT300) and TIC (FLIR E8, Model: E6390). The Root Mean Square metric was applied to obtain a measure of the degree of fitness. The model fit was validated using several metrics, with the Root Mean Square Error metric. TIC readings were 1.0 to 1.5oC higher when compared with IRT readings (P<0.0001). The back temperature registered using IRT was higher in male rats when compared with female rats (P<0.0001). In mice, the tail temperature measured using IRT was slightly higher in females when compared with males (P<0.0001). We found that it is possible to fit, by linear regression, a straight line to the data, to obtain a linear transformation that allows inferring the TIC readings from the IRT readings, obviating the use of TIC readings.