To improve the infrared thermal imaging screening method and compare the skin temperature difference between the affected knee and the contralateral knee before the total knee arthroplasty(TKA), and to establish the pattern of knee skin temperature before knee surgery that may be significant for early auxiliary evaluation of knee joint osteoarthritis.

This study excluded the knee joint osteoarthritis patients which had incomplete preoperative evaluation information and were treated with surgery. Thirty-one patients who met the criteria of knee joint osteoarthritis diagnosis were included in the study; their knee joint skin temperature images were captured one day before TKA by infrared thermography. This study adopted the method of comparing the affected knee and the contralateral knee to calculate the temperature difference and analyzed a new partition of the anterior aspect and the side face of knee. The t test was applied to the normal distribution data, while the rank sum test was used to the nonnormal distribution data.

There was statistically significant in every region mean temperature difference(MTD) between the affected side and contralateral side in the anterior aspect (P<0.05). Among them, the MTD regions of medial, patella central, inferomedial, inferior were highly statistically significant (t=2.992, 4.188, 4.624, 4.127; all P<0.01). The most different region of MTD was the inferomedial. In the lateral side, it has no statistic significance of MTD. In the medial side, the central-knee and distal-knee region MTD were statistically significant(t=1.867, 0.887; both P<0.05). The most different region of MTD was the distal-knee. Besides, the percentages of the area between the affected side and contralateral side were statistically significant different in the temperature range above 31℃, 32 ℃, 33 ℃, 34 ℃, 35 ℃ in the anterior aspect (t=3.097, 4.231, 4.621, 4.574, Z=3.774; all P<0.01). And there was statistically significant difference in the temperature range above 30℃(t=2.311, P <0.05). The percentages of the area between the affected side and contralateral side was no statistically significant different above the node from 28 ℃ to 36 ℃ in the lateral side and the medial side.

The affected knee skin temperature of patients with knee osteoarthritis is higher than the contralateral side of the knee and the maximum difference region of MTD may be the inferomedial of the affected knee. There is an obvious difference in the anterior aspect of the knee joint temperature area distribution. The max value of area ratio difference is in the temperature range above 33℃. These research results may provide an idea to early assisted assessment knee osteoarthritis with suitable new type of thermal imaging technology.