Critical Review of Generative Adversarial Networks in Dermatological Applications with Focus on Nail Disease Detection

Generative Adversarial Networks, nail disease detection, onychomycosis, dermatology, medical image synthesis, data augmentation, conditional GAN, systematic review.

Dr.Khel Prakash Jayant,
Professor, Department of Computer Science & Engineering,
Raj Kumar Goel Institute of Technology, Uttar
Pradesh, India. Email kpjayant@gmail.com.
Orcid ID- 0000-0001-8855-8559

Dr. Pramod Kumar Sagar,
Associate Professor, Department of Computer Science &
Engineering, Raj Kumar Goel Institute of
Technology, Uttar Pradesh, India. Email:
pksagar1975@gmail.com.

Dr Amit Asthana,
Associate professor, Computer Science and Engineering,
SGT University Gurgaon.
Email: amitasthana_soet@sgtuniversity.org

Mukul Maurya,
IIMT University, Meerut,
India 250001,
Email:mukulmaurya14@gmail.com

Dr. Pushpendra Kumar Verma,
IIMT University, Meerut, India 250001,
Email: dr.pkverma81@gmail.com

Amit Kumar Sharma,
IIMT University,
Meerut, India 250001,
Email: fspeed0007@gmail.com

Dr. Madhu Yadav,
IIMT University, Meerut,
India 250001.
Email:my8006069850@gmail.com

Received:14-06-2026
Revised:17-07-2026
Accepted: 25-07-2026

European Journal of Prosthodontics and Restorative Dentistry (2026) 34(7s), 680–701

Critical Review of Generative Adversarial Networks in Dermatological Applications with Focus on Nail Disease Detection

Generative Adversarial Networks (GANs) have become a groundbreaking technology in medical image analysis, enabling realistic synthetic image generation, data augmentation, and improved diagnostic results. GANs have been used in dermatology to classify skin lesions, analyze pigmentation, and detect melanoma. Nonetheless, the field of nail disease diagnosis that includes onychomycosis, nail psoriasis, and subungual melanoma is highly under-investigated despite the high clinical impact and distinct diagnostic features of nail structure. Comparative analysis of DCGAN, conditional GAN (cGAN), cycleGAN and StyleGAN shows that although GANs have a great impact on improving classification and reducing the problem of class imbalance in skin lesion analysis, with an increase in accuracy of 5-12% in minority classes, there is almost no attempt to apply GANs to nail diseases. Few studies discuss nail conditions and none of them proposes a complete GAN-based diagnostic system. Among the key challenges, training instability, mode collapse, no labeled nail data, ethical and privacy concerns, and no clinical validation have been noted. Future research opportunities involve creating large and publicly accessible nail datasets, hybrid GANtransformer models, explainable AI in clinical trust, and implementation in mobile healthcare. This review finds that GANs have a huge untapped potential of nail disease detection, and that existing methods of investigating skin lesions may be transferred to nail imaging to fill critical gaps in nail imaging, eventually enhancing diagnostic accuracy and patient outcomes.

•••••••••••••••••••••••••••••••• Ejprd.org - Published by Riset Publication Services LLC

EJPRD

Copyright ©2026 by Riset Publication Services LLC