Infrared thermography (IRT) – a non-contact, non-invasive technique – has been used for mass screenings to identify febrile individuals at transportation nodes (e.g., airports) during infectious disease pandemics such as SARS (Severe Acute Respiratory Syndrome), H1N1 virus, and Ebola outbreaks. Despite the potential of IRTs, the field lacks a well-established consensus methodology to ensure temperature measurement accuracy and reliability. This study aims to investigate the use of IRTs in a controlled setting to determine the effectiveness of IRT and the most reliable facial region for estimation of core temperature. We conducted a large clinical study, acquiring facial thermographs of 1,109 febrile and non-febrile subjects using Screening Thermographs (STs). Regression analyses between the reference oral temperature and different areas of the face, specifically the forehead and canthi, were carried out. The coefficients of determination of each regression were compared to determine how well facial and core body temperatures were correlated. Receiver operating characteristic (ROC) curves were constructed to compare the effectiveness of using different facial areas to identify febrile patients. Results show that the maximum temperature of the overall face has the best linear trend, followed by the maximum temperature at the inner canthus region. Both of these values show better correlations than forehead temperatures, which are commonly used as a target by non-contact infrared thermometers. For any chosen facial area, the maximum temperature collected always showed a stronger correlation than a specific point in that area. Results indicate that IRT performance is substantially approved when applying optimal measurement methodology.
|