Smartphone-based mobile thermal imaging technology to assess limb perfusion and tourniquet effectiveness under normal and blackout conditions

BACKGROUND Over the past decade, there has been a resurgence of tourniquet use in civilian and military settings. Several key challenges include assessment of limb perfusion and adequacy of tourniquet placement, particularly in the austere or prehospital environments. We investigated the utility of thermal imaging to assess adequacy of tourniquet placement. METHODS The FLIR ONE smartphone-based thermal imager was utilized. Ten swine underwent tourniquet placement with no associated hemorrhage (n = 5) or with 40% hemorrhage (n = 5). Experiment 1 simulated proper tourniquet application, experiment 2 had one of two tourniquets inadequately tightened, and experiment 3 had one of two tourniquets inadequately tightened while simulating blackout-combat conditions. Static images were taken at multiple time points up to 30 minutes. Thermal images were then presented to blinded evaluators who assessed adequacy of tourniquet placement. RESULTS The mean core temperature was 38.3 °C in non-hemorrhaged animals versus 38.2 °C in hemorrhaged animals. Hemorrhaged animals were more hypotensive (p = 0.001), anemic (p < 0.001), vasodilated (p = 0.008), and had a lower cardiac output (p = 0.007) compared to non-hemorrhaged animals. The thermal imaging temperature reading decreased significantly after proper tourniquet placement in all animals, with no difference between hemorrhaged and non-hemorrhaged groups at 30 minutes (p = 0.23). Qualitative thermal image analysis showed clearly visible perfusion differences in all animals between baseline, adequate tourniquet, and inadequate tourniquet in both hemorrhaged and non-hemorrhaged groups. Ninety-eight percent of blinded evaluators (n = 62) correctly identified adequate and inadequate tourniquet placement at 5 minutes. Images in blackout conditions showed no adverse impact on thermal measurements or in the ability to accurately characterize perfusion and tourniquet adequacy. CONCLUSIONS A simple handheld smartphone-based forward looking infrared radiometry device demonstrated a high degree of accuracy, reliability, and ease of use for assessing limb perfusion. Forward looking infrared radiometry also allowed for rapid and reliable identification of adequate tourniquet placement that was not affected by major hemorrhage or blackout conditions.