Forward-looking infrared imaging predicts ultimate burn depth in a porcine vertical injury progression model

Introduction: Current methods of assessing burn depth are limited and are primarily based on visual assessments by burn surgeons. This technique has been shown to have only 60% accuracy and a more accurate, simple, noninvasive method is needed to determine burn wound depth. Forward-looking infrared (FLIR) thermography is both noninvasive and user-friendly with the potential to rapidly assess burn depth. The purpose of this paper is to determine if early changes in burn temperature (first 3 days) can be a predictor of burn depth as assessed by vertical scarring 28 days after injury.

Methods: While under general anesthesia, 20 burns were created on the backs of two female Yorkshire swine using a 2.5cm×2.5cm×7.5cm, 150g aluminum bar, for a total of 40 burns. FLIR imaging was performed at both early (1, 2 and 3 days) and late (7, 10, 14, 17, 21, 24 and 28 days) time points. Burns were imaged from a height of 12 inches from the skin surface. FLIR ExaminIR(©) software was used to examine the infrared thermographs. One hundred temperature points from burn edge to edge across the center of the burn were collected for each burn at all time points and were exported as a comma-separated values (CSV) file. The CSV file was processed and analyzed using a MATLAB program. The temperature profiles through the center of the burns generated parabola-like curves. The lowest temperature (temperature minimum) and a line midway between the temperature minimum and ambient skin temperature at the burn edges was defined and the area of the curve calculated (the "temperature half-area").

Results: Half-area values 2 days after burn had higher correlations with scar depth than did the minimum temperatures. However, burns that became warmer from 1 day to 2 days after injury had a lower scar depth then burns that became cooler and this trend was best predicted by temperature minima. When data were analyzed as a diagnostic test for sensitivity and specificity using >3mm scarring, i.e. a full-thickness burn, as a clinically relevant criterion standard, temperature minima at 2 days after burn was found to be the most sensitive and specific test.

Conclusions: FLIR imaging is a fast and simple tool that has been shown to predict burn wound outcome in a porcine vertical injury progression model. Data showed that more severe burn wounds get cooler between 1 and 2 days after burn. We found four analytic methods of FLIR images that were predictive of burn progression at 1 and 2 days after burn; however, temperature minima 2 days after burn appeared to be the best predictive test for injury progression to a full-thickness burn. Although these results must be validated in clinical studies, FLIR imaging has the potential to aid clinicians in assessing burn severity and thereby assisting in burn wound management.