Initial assessment of burn injuries is critical to predicting optimal treatment, but such inspection alone has shown to be inaccurate in 30-50% of cases. Burn wound analysis and implementation of appropriate remedial therapies are improved by use of a non-invasive digital infrared thermography technique developed by researchers from Universidad Autónoma de San Luis Potosí, Mexico, and McGill University, Canada.

The protocol relies on digital thermograms obtained with a thermal camera to determine thermal properties of Clinical images (A to C) and thermograms (A1 to C1) were obtained during the first three days after the injury, the patients were followed up until discharged, and their outcome classified as healed by conservative treatment, skin graft or amputation. Source: Universidad Autónoma de San Luis Potosí, McGill Universitythe wound. Differences in the average temperature of the burn and of an area of adjacent healthy skin are then recorded.

After patients were treated and released as they normally would be, the researchers categorized the treatment given as re-epithelization if the wound healed by itself, as skin graft if the patient received at least one or as amputation if any extremity had to be removed due to extensive damage. A prediction model was formulated with a machine learning method to categorize treatment based on the clinical data and the thermograms, which revealed that the temperature difference between healthy skin and the wound alone is enough to make the prediction. The accuracy of the prediction using their algorithm was found to be 90% both in the development cohort of patients, as well as in an independent validation cohort.

The method could prove useful for early assessment of patients in the emergency department, as well as in later stages of patient care to identify dead tissue and infections or other complications. A research paper on the digital infrared thermography trials is published in PLOS ONE.