Metraipolis › archive › dc_319148b0bf77

Disease Burden over Skin Tone: Decomposing the Dermatology-AI Generalization Gap — arXiv:2609.02111

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.02111
License (as archived)
CC0-1.0
Archived
2026-10-17T03:37:09.162Z
SHA-256 of the archived text
c82b404fdb2839a48e0c85753962a07afc5aeee0eb34c15b59e316133197ab36
Stable URL of this page
https://metraipolis.com/documento/dc_319148b0bf77

How to cite

Disease Burden over Skin Tone: Decomposing the Dermatology-AI Generalization Gap — arXiv:2609.02111 — arXiv (en), https://arxiv.org/abs/2609.02111 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_319148b0bf77 (Metraipolis archive dc_319148b0bf77, sha256 c82b404fdb2839a48e0c85753962a07afc5aeee0eb34c15b59e316133197ab36)
BibTeX
@misc{dc319148b0bf77,
  title = {Disease Burden over Skin Tone: Decomposing the Dermatology-AI Generalization Gap — arXiv:2609.02111},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.02111},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_319148b0bf77; sha256 c82b404fdb2839a48e0c85753962a07afc5aeee0eb34c15b59e316133197ab36},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.02111. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Dermatology artificial intelligence (AI) models are predominantly trained on light-skinned, cancer-focused image collections, yet they are increasingly proposed for deployment in resource-constrained settings where patients differ from training populations along two confounded axes: skin tone and disease distribution. We investigate whether poor generalization is primarily caused by skin-tone underrepresentation or disease-distribution shift. We evaluate a cancer-trained baseline (ResNet-50 fine-tuned on HAM10000 and ISIC 2019), two dermatology foundation models (DermLIP and MONET), and a general-purpose vision model (DINOv3) as frozen feature extractors. Models are evaluated on a tone-stratified disease-matched dataset (Diverse Dermatology Images, DDI) and a disease-shifted tone-diverse dataset (Skin Condition Image Network, SCIN). Our results show that disease-distribution shift contributes more than skin tone in the evaluated settings. The cancer baseline decreases from 0.62 to 0.21 balanced accuracy when transferred to unfamiliar clinical conditions, while the within-disease skin-tone gap is smaller (0.10-0.18) and inconsistent. Label-free representation analysis shows that this failure reflects a representational limitation rather than only missing output labels: cancer-specialized features poorly cluster unfamiliar conditions (kNN purity lift +0.06 over chance), whereas dermatology-pretrained features retain stronger transferable structure (+0.23). Finally, we show that representation quality predicts recoverable performance under lightweight adaptation. Starting from dermatology foundation models, approximately ten labeled examples per clinical category recover most attainable performance. We release the evaluation protocol and code to support reproducible auditing of dermatology AI generalization.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-03 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.02111 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T03:13:12.505Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-03?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).