Distance and camera features measurements affect the detection of temperature asymmetries using infrared thermography

Abstract

ABSTRACT Small skin temperature asymmetries are claimed to indicate injuries, but the influence of measurement conditions and technical aspects of image acquisition are concerns for determining temperature asymmetries. This research determines whether distance and camera specifications affect thermal asymmetries detection. We simulated thermal asymmetries filling two glasses with water at different temperatures measured at two distances (0.7 and 1.5 m) using four different infrared cameras (T1020, E60BX, C2 and FLIR ONE). Linear regression verified the similarity between each camera and the T1020 camera (gold standard). The intraclass correlation coefficient assessed the inter-camera reproducibility. Linear regression reported increasing adjustment using mean temperature, with lower values at 1.5 m between the T1020 and FLIR ONE (0.7 m, r = 0.58; 1.5 m, r = 0.34), C2 (0.7 m, r = 0.90; 1.5 m, r = 0.62) and E60BX (0.7 m, r = 0.96; 1.5 m, r = 0.68). Higher inter-camera reproducibility was observed at 0.7 m than 1.5 m. It is preferable to take thermal images with shorter distances and to prioritise cameras with superior hardware characteristics. The portability of FLIR ONE and C2 may be an advantage in studies investigating expected differences greater than 0.05ºC and larger regions of interest.