ABSTRACT
The spectrum of allergic diseases includes atopic dermatitis (AD), allergic rhinitis (AR), and asthma. To date, the association between allergic diseases and psoriasis has not yet been completely evaluated. This study was conducted to determine the risk of psoriasis in patients with allergic diseases. A health screening database, a sub-dataset of the Korean National Health Insurance Service database, was used. All 9,718,722 subjects who underwent health examination in 2009 at age over 20 were included. Subjects with allergic diseases including AD (n = 35,685), AR (n = 1,362,713), asthma (n = 279,451) and control subjects without all three allergic diseases (n = 8,210,042), without AD (n = 9,683,037), without AR (n = 8,356,009) and without asthma group (n = 9,439,271) were analyzed. The subjects were tracked using their medical records during the 8-year period from 2010 to 2017 to identify those who developed psoriasis. Multivariate Cox regression models were used to assess the risk of psoriasis. The incidence probability of psoriasis was analyzed through the Kaplan-Meier method. The incidence of psoriasis per 1,000 person-years was 9.57, 3.78, and 4.28 in the AD, AR, and asthma groups, respectively. The AD group exhibited a significantly increased risk of developing psoriasis compared to subjects without AD (hazard ratio [HR], 3.18; 95% confidence interval [95% CI], 3.05-3.31; P < 0.001) after adjustment for confounding factors. The risk of psoriasis was significantly increased in the AR group compared to subjects without AR (HR, 1.32; 95% CI, 1.31-1.34; P < 0.001) and asthma group compared to subjects without asthma (HR, 1.30; 95% CI, 1.27-1.33; P < 0.001). Allergic diseases, particularly AD, may be a risk factor for psoriasis.
ABSTRACT
Scoring atopic dermatitis (AD) severity with the Eczema Area and Severity Index (EASI) in an objective and reproducible manner is challenging. Automated measurement of erythema, papulation, excoriation, and lichenification severity using images has not yet been investigated. Our aim was to determine whether convolutional neural networks (CNNs) could assess erythema, papulation, excoriation, and lichenification severity at a level of competence comparable to dermatologists. We created a standard dataset of 8,000 clinical images showing AD. Each component of the EASI was scored from 0 to 3 by three dermatologists. We trained four CNNs (ResNet V1, ResNet V2, GoogLeNet, and VGG-Net) with the image dataset and determined which CNN was the most suitable for erythema, papulation, excoriation, and lichenification scoring. The brightness of the images in each dataset was adjusted to - 80% to + 80% of the original brightness (i.e., 9 levels by 20%) to investigate if the CNNs accurately measured scores if image brightness levels were changed. Compared to the dermatologists' scoring, accuracy rates of the CNNs were 99.17% for erythema, 93.17% for papulation, 96.00% for excoriation, and 97.17% for lichenification. CNNs trained with brightness-adjusted images achieved a high accuracy without the need to standardize camera settings. These results suggested that CNNs perform at level of competence comparable to dermatologists for scoring erythema, papulation, excoriation, and lichenification severity.
