Faculty Sponsor

Jana Stedman

College

College of Nursing & Health Professions (CONHP)

Department/Program

Master of Science in Physician Assistant

Presentation Type

Poster Presentation

Symposium Date

Summer 7-22-2026

Abstract

Abstract

Objective: Melanoma is the deadliest form of skin cancer, and early detection significantly improves survival. The purpose of this literature review was to evaluate the diagnostic accuracy of artificial intelligence (AI)-assisted dermoscopy compared with clinician assessment for melanoma detection among adults with suspicious pigmented skin lesions. Biopsy-confirmed diagnosis served as the reference standard.

Methods: A comprehensive literature search was conducted using PubMed, EBSCO, Research Rabbit, and Google Scholar. Studies published from 2020 to 2025 involving adults with suspicious pigmented skin lesions were included if they compared AI-assisted dermoscopic analysis with clinician interpretation using biopsy-confirmed diagnosis and reported sensitivity and specificity. Randomized controlled trials, cohort studies, cross-sectional studies, systematic reviews, and meta-analyses were reviewed.

Results: Across the five included studies, AI-assisted dermoscopy demonstrated higher sensitivity for melanoma detection than clinician assessment, particularly among nonspecialist and less experienced providers. Several studies reported that AI diagnostic performance was comparable to expert dermatologists; however, AI achieved higher sensitivity at the expense of lower specificity, resulting in increased false-positive findings.

Conclusion: Current evidence supports AI-assisted dermoscopy as a valuable clinical decision support tool for melanoma detection rather than a replacement for clinician judgment. AI appears most beneficial for improving early melanoma detection, particularly in nonspecialist settings. Limitations in specificity, real-world validation, and representation of diverse patient populations highlight the need for further research before widespread implementation.

Keywords: artificial intelligence; dermoscopy; melanoma; sensitivity; specificity; diagnostic accuracy.

Biographical Information about Author(s)

Sandra Marinceski has an anticipated graduation date of July 2026 from the Valparaiso University Physician Assistant Program. Throughout her clinical year, she enjoyed exploring a variety of specialties, with particular interests in dermatology and emergency medicine. One of her favorite aspects of clinical rotations was building relationships with patients while gaining hands on experience across diverse healthcare settings. Sandra looks forward to beginning her career as a physician assistant and continuing to grow as a healthcare provider. She was inspired to research the use of artificial intelligence in melanoma detection due to her interest in dermatology and emerging technologies that have the potential to improve patient care. 

Download

Share

COinS