Developing an Algorithm to Quantify Dermatomyositis-Associated Calcinosis Using Infrared Images

Background/Objectives: Adult and Juvenile Dermatomyositis (DM/JDM)-associated calcinosis varies in its clinical presentations. Current modalities such as physician exam and CT imaging do not fully document calcinosis burden. As a result, this paper examines infrared imaging, which measures skin temperature, as a supplementary modality to assess calcinosis activity. Methods: Fifteen patients (3 DM and 12 JDM) who fulfilled probable or definite Bohan and Peter criteria for DM/JDM and with previously documented calcinosis were included. IR scans of target body regions were obtained in a temperature-controlled and humidity-monitored room. We developed an algorithm to assess calcinosis by measuring local temperature variations in IR scans. Body regions were determined to be calcinosis-positive if the algorithm identified at least 15% of the body region as suspected calcinosis. Results: Twenty-nine out of ninety body regions were designated calcinosis-positive. The average suspected calcinosis areas ranged from 8.6 to 14.2%, and left and right extremity averages were within 0.6% area of each other. Agreement between the IR algorithm and physician exam was variable, as the algorithm simultaneously detected regions that overlapped with physician-marked areas and regions outside of physician-marked boundaries. Conclusions: IR imaging appears to identify areas of calcinosis activity in DM/JDM and presents many promising advantages to warrant future study. Its integration with structural imaging modalities will enable a more comprehensive assessment of calcinosis.