• facial erythema;
  • rosacea;
  • objective assessment;
  • computer-aided image analysis;
  • color deconvolution


Facial erythema is a common but morbid condition caused by several inflammatory disorders. Since most of the current severity indexes lack objectivity and global assessment, the computer-aided image analysis (CAIA) has been suggested.


The purpose of this study is to compare the validity and efficacy of novel CAIA erythema index – erythema dose (ED) with two other CAIA indexes.


For mild, moderated and severe erytherotelangiectatic rosacea patients, their photographs of right cheek were assessed with red-blue difference index in image (RBI), a*, and ED. For each index, the differences between severity groups were analyzed. The correlations between the three indexes were evaluated. To evaluate the capability of differentiating the pathologic erythema from the red component of the normal skin, each index was calculated at both the erythema and normal area.


The a* and ED significantly changed according to the severity, while the RBI did not. The ED was different between both mild-moderate and moderate-severe, while the a* only between mild-moderate. There was a strong positive correlation between a* and ED. The difference between erythema and normal skin was more prominent in ED than in a*.


We confirmed that the ED is well correlated with both subjective assessment, as well as the other CAIA index a*. The ED has the advantage of specifically visualizing and analyzing the pathologic erythema.