A Multicentric Case–Control Study for Diagnostic Utility of Non-contact Infrared Thermography (IRT) in Type 2 Diabetes

Abstract

While there is a renewed interest in the use of infrared imaging in the medical field owing to improvements in image acquisition and processing technologies, its utility in the diagnosis of diabetes mellitus on a large scale has not been adequately evaluated in previous studies. This was a multicenter, double-blind case–control trial across three centers in India which enrolled a total of 1095 adults in two groups—590 participants with diabetes and 505 healthy controls. The primary outcome was the difference in thermography readings between the two groups. The secondary outcomes included the correlation of thermography readings with HbA1C, considered the gold standard. The thermography readings at all seven sites were lower in people with diabetes as compared to the control group. However, this attained statistical significance across four sites—forehead (Δmean difference = 0·34, 95% CI [0·14–0·53], medial canthus of both eyes (Δmean difference = 0·48, 95% CI [0·31–0·66], pinna of both ears (Δmean difference = 0·53, 95% CI [0·28–0·73]) and palmar surface of both hands (Δmean difference = 0·28, 95% CI [0·02–0·53]). In addition, thermography readings showed a significant negative correlation with HbA1C values across these four sites, although the strength of the correlation was weak. This study sheds light on the utility of thermography for large-scale noninvasive diagnosis of diabetes mellitus. Future research should focus on large-scale community screening of early asymptomatic disease as well as micro- and macrovascular complications to define its exact role in the overall disease management pathway.