ARTIFICIAL INTELLIGENCE IN DERMATOLOGY: CURRENT STATE, CHALLENGES AND FUTURE PERSPECTIVES
DOI:
https://doi.org/10.52340/jecm.2025.06.30Keywords:
AI, Dermatology, Dermoscopy, Skin Cancer, Melanoma, Deep Learning, CNNAbstract
Artificial intelligence has become a crucial tool in modern dermatology, enhancing diagnostic accuracy for skin neoplasms, reducing overdiagnosis, and providing standardized assessments. CNN-based models perform at or above dermatologist level in detecting melanoma and other malignancies (AUC > 0.94). AI platforms such as PhotoFinder and Doctorium help decrease unnecessary biopsies by 20–30%, significantly enhances diagnostic yield, achieving a validated diagnostic accuracy of 90% or greater. AI is widely applied in dermoscopy, histopathology, and trichoscopy, enabling automatic structure recognition, mitotic count analysis, and classification of hair loss disorders. Despite its significant potential, challenges remain, including data heterogeneity, algorithm transparency, and ethical considerations. The future of AI in dermatology lies in multimodal models, real-time risk prediction, and the integration of locally developed Georgian AI systems into clinical practice.
Downloads
References
Brinker T.J., Hekler A., Enk A., et al. Deep neural networks are superior to dermatologists in melanoma image classification. European Journal of Cancer. 2019;119:11–17.
Codella N., Rotemberg V., Tschandl P., et al. Skin lesion analysis toward melanoma detection: A challenge hosted by the International Skin Imaging Collaboration (ISIC) 2018. arXiv preprint arXiv:1902.03368.
Daneshjou R., Vodrahalli K., Novoa R., et al. Disparities in dermatology AI: assessing algorithm performance across skin tones. Science Advances. 2022;8(15):eabq6147.
Doctorium AI Platform. Algorithmic Framework for Skin Lesion Risk Stratification, 2024
Esteva A., Kuprel B., Novoa R., et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature. 2017;542:115–118.
Haenssle H.A., Fink C., Schneider S., et al. Man against machine: diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition. Annals of Oncology. 2018;29(8):1836–1842.
ISIC Archive – International Skin Imaging Collaboration. 2016–2024. https://isic-archive.com
Kittler H., Pehamberger H., Wolff K., Binder M. Diagnostic accuracy of dermoscopy. The Lancet Oncology. 2002;3(3):159–165.
Liu Y., Jain A., Eng C., et al. A deep learning system for differential diagnosis of skin diseases. Nature Medicine. 2020;26:900–908.
PhotoFinder Systems. Artificial Intelligence in Dermoscopy – Technical Documentation, 2022.
Vanderbilt University Medical Center. Whole Slide Imaging and Artificial Intelligence in Pathology. VUMC Pathology Review, 2021.
Downloads
Published
How to Cite
Issue
Section
