EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Adversarial robustness of dental panoramic radiograph classifiers against gradient-based perturbations

Wenya Long, Teng He, Yingjuan Xie

SAP Artificial Intelligence in Dentistry · 01.01.2027 · 3 · 186

DOI: 10.62486/aid2027186 ↗

Kurz gefasst

Fachartikel in der Zeitschrift SAP Artificial Intelligence in Dentistry, online erschienen am 01.01.2027. Verfasst von Wenya Long und 2 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Deep learning classifiers for dental radiographs are increasingly deployed, but their robustness to adversarial perturbations is unmeasured.

Methods

A frozen ResNet-18 backbone and a linear head were trained on 182 annotated teeth from the DENTEX dataset and attacked with the fast gradient sign method and projected gradient descent at budgets from 1/255 to 16/255, comparing a baseline model, an adversarially trained model and a median-filter defense.

Results

The baseline classified 63.6% of held-out teeth correctly. Both attacks flipped every correctly classified tooth at a budget of 1/255. The fast gradient sign method fell to 77.1% attack success at 16/255, whereas projected gradient descent remained at 100%. Neither adversarial training nor the median filter reduced attack success, and adversarial training lowered clean accuracy to 54.5%.

Conclusions

A transfer-learning dental radiograph classifier with a frozen backbone and linear head is catastrophically vulnerable to adversarial perturbations, and the simple defenses tested do not restore robustness.

Zitieren

APA

Long, W., He, T., & Xie, Y. (2027). Adversarial robustness of dental panoramic radiograph classifiers against gradient-based perturbations. SAP Artificial Intelligence in Dentistry, 3, 186. https://doi.org/10.62486/aid2027186

Vancouver

Long W, He T, Xie Y. Adversarial robustness of dental panoramic radiograph classifiers against gradient-based perturbations. SAP Artificial Intelligence in Dentistry. 2027;3:186. doi:10.62486/aid2027186

BibTeX
@article{Long2027,
  author = {Long, Wenya and He, Teng and Xie, Yingjuan},
  title = {Adversarial robustness of dental panoramic radiograph classifiers against gradient-based perturbations},
  journal = {SAP Artificial Intelligence in Dentistry},
  year = {2027},
  volume = {3},
  pages = {186},
  doi = {10.62486/aid2027186}
}
Kostenlos starten →