Infrared Thermal Imaging to Detect Inflammatory Intra-Abdominal Pathology in Infants

Abstract

A thermal imaging method to detect inflammatory intra-abdominal pathology in infants is proposed and evaluated through a clinical trial. Nine surgical infants, mean chronological age 58 days old (range: 21-83 days), mean weight 2.65 kg (range: 2.45-3.15 kg) with abdominal pathologies were included in the analysis. Infrared thermal image processing consisted of selecting the surgical region of interest where the area of abdominal inflammation was most likely to be, and an abdominal reference region on the same infant, with the aid of clustering segmentation. Skewness was found to be the most sensitive variable to significantly differentiate between the surgical region and reference region (p = 0.022). Multilinear regression analysis indicated that the relationship between the temperature difference signified by skewness and the patients' demographic information (age at time of imaging, gestational age at birth, weight at the time of imaging, birthweight, last stool prior to imaging and last oral intake prior to imaging) was not significant. The study indicated that inflammatory regions, such as those found in infants following surgery, would have a significantly different temperature distribution than the surrounding skin. The method differentiated between an inflammatory and non-inflammatory region on the abdomen