EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Facial-Relevant Dermatology Image Datasets for AI: Biases, Benchmarking, and Readiness for Vision-Language Models

Rumeysa Gulcihan Cakmak, Buse Karaman, Zeynep Beceren, Duygu Cakir

IEEE Access · 2026 · 14 · 164469-164491

DOI: 10.1109/access.2026.3732560 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Rumeysa Gulcihan Cakmak und 3 weiteren Autor:innen. Die Arbeit nennt 102 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Artificial intelligence (AI) systems for dermatology increasingly rely on heterogeneous image resources, yet facial images combine diagnostic value with biometric identifiability, acquisition variability, and psychosocial sensitivity. Searches through March 2026 identified 60 candidate datasets or resources. After source, relevance, accessibility, and overlap screening, 25 core entries were retained for structured comparison; 2 additional resources were discussed for taxonomy or emerging-method context, and 33 remained supplementary-only. Among the 25 core entries, only 3 are explicitly facial-specific, 12 are broad clinical resources in which facial involvement is plausible but not guaranteed, and 10 are dermoscopic or total-body-photography benchmarks with indirect facial relevance. Only the three facial-specific entries are treated as direct evidence about facial-image datasets; the remaining entries provide broader clinical or benchmarking context. Fourteen entries have peer-reviewed primary or official challenge evidence, three are official-atlas or emerging resources, and eight are community or derivative sources requiring stronger provenance checks. We compare scale, modality, annotation, metadata, licensing, demographic reporting, task suitability, and benchmarking practice, and assess readiness for vision-language model development. Persistent limitations include documented demographic imbalance in some resources and, more broadly, sparse or inconsistent demographic and skin-tone reporting, incomplete body-site metadata, inconsistent taxonomies, weak external validation, and inappropriate cross-task metric comparison. We conclude with recommendations for modality-specific benchmarking, clinician review, privacy-preserving governance, and clinically validated multimodal annotation.

Zitieren

APA

Cakmak, R. G., Karaman, B., Beceren, Z., & Cakir, D. (2026). Facial-Relevant Dermatology Image Datasets for AI: Biases, Benchmarking, and Readiness for Vision-Language Models. IEEE Access, 14, 164469-164491. https://doi.org/10.1109/access.2026.3732560

Vancouver

Cakmak RG, Karaman B, Beceren Z, Cakir D. Facial-Relevant Dermatology Image Datasets for AI: Biases, Benchmarking, and Readiness for Vision-Language Models. IEEE Access. 2026;14:164469-164491. doi:10.1109/access.2026.3732560

BibTeX
@article{Cakmak2026,
  author = {Cakmak, Rumeysa Gulcihan and Karaman, Buse and Beceren, Zeynep and Cakir, Duygu},
  title = {Facial-Relevant Dermatology Image Datasets for AI: Biases, Benchmarking, and Readiness for Vision-Language Models},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {164469--164491},
  doi = {10.1109/access.2026.3732560}
}
Kostenlos starten →