Abstract
Joint inflammation at the bilateral knees of patients with rheumatoid arthritis (RA) was assessed using thermal imaging (TI) in comparison with ultrasonography. The minimum (T-min), maximum (T-max) and average (T-avg) temperatures at the lateral, anterior and medial knee aspects were correlated with ultrasound-detected grey-scale (GS) and power Doppler (PD) joint inflammation. The use of TI in identifying ultrasound GS score ≥ 2 and PD score > 0 were assessed using receiving operating characteristic (ROC) curve analysis. A subset of baseline thermograms were re-read > 2 weeks apart for intra-observer reliability testing using intra-class correlation coefficient (ICC) analysis. In this cross-sectional study (n = 95 RA patients), 570 thermograms were obtained while 190 knees were scanned by ultrasound. All TI parameters (T-min, T-max and T-avg) correlated significantly (P < 0.05) with ultrasound PD and GS scores (correlation coefficient ranged from 0.21 to 0.49). Using TI parameters to identify ultrasound PD score > 0 and GS score ≥ 2, the area under the ROC curves (AUCs) ranged from 0.63 to 0.82. The ICC values for the TI parameters were high (ranging from 0.997 to 0.999). Thermographic heat signatures are reflective of ultrasound-detected joint inflammation and can help discriminate ultrasound PD positivity/negativity and GS joint inflammation severity at the knees in patients with RA.