A New Deep Learning Model Discriminates Between Scalp Psoriasis and Seborrheic Dermatitis

A new deep learning-based approach shows promising results in differentiating between scalp psoriasis and seborrheic dermatitis from dermoscopic images, according to a study published in Frontiers in Medicine.

Researchers aimed to develop a deep learning approach for differentiating scalp psoriasis and seborrheic dermatitis from dermoscopic images that would achieve a higher accuracy than dermatologists trained with dermoscopy. A total of 617 patients with pathological and diagnostic confirmed skin diseases were used to collect 1358 pictures. The pictures were randomly divided into training, validation, and testing datasets in the study.

The deep learning model showed 96.1% sensitivity, 88.2% specificity, and 0.922% area under curve. It outperformed all dermatologists in diagnosing scalp psoriasis and seborrheic dermatitis. It also helped a dermatology graduate study and general practitioners improve their diagnostic performance.

“The developed [deep learning] model has favorable performance in discriminating two skin diseases and can improve the diagnosis, clinical decision-making, and treatment of dermatologists in primary hospitals,” concluded the study authors.

Reference
Yu Z, Kaizhi S, Jianwen H, Guanyu Y, Yonggang W. A deep learning-based approach toward differentiating scalp psoriasis and seborrheic dermatitis from dermoscopic images. Front Med (Lausanne). Published online November 3, 2022. doi:10.3389/fmed.2022.965423