Artificial intelligence on inflammatory dermatoses: where we are and where are we going?

Area of Science:

• Dermatology
• Artificial Intelligence
• Medical Imaging Analysis

Background:

• Artificial intelligence (AI) demonstrates diagnostic accuracy comparable to dermatologists for skin lesions.
• Current AI development in dermatology primarily targets neoplastic diseases due to high prevalence and morbidity.
• The focus is shifting towards exploring AI's potential in other dermatological areas.

Purpose of the Study:

• To review and describe the emerging applications of artificial intelligence in the field of inflammatory dermatoses.
• To consolidate current research on AI's role in managing non-neoplastic skin conditions.

Main Methods:

• A systematic literature search was conducted in Medline and Lilacs databases.
• Articles published between 2013 and 2023 were screened using specific inclusion and exclusion criteria.
• Information from 19 selected articles was extracted to synthesize the review findings.

Main Results:

• Early AI dermatology research concentrated on neoplasms like melanoma, achieving high diagnostic accuracy.
• Recent studies increasingly explore AI applications in inflammatory dermatoses, including psoriasis, hidradenitis suppurativa, and atopic dermatitis.
• AI tools are being developed to assist in objective assessments, such as calculating Psoriasis Area and Severity Index (PASI) scores.

Conclusions:

• The study is a literature review, acknowledging the innovative nature and limited published research in AI for inflammatory dermatoses.
• Despite limitations, a growing interest in AI for dermatology is evident, with significant potential for future advancements.
• Artificial intelligence is poised to become a valuable tool to support dermatologists in clinical practice.