EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Bias Detection in Histology Images Using Explainable AI and Image Darkness Assessment

Inna Skarga-Bandurova, Golshid Sharifnia, Tetiana Biloborodova

Studies in Health Technology and Informatics · 2025

DOI: 10.3233/shti250302 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Studies in Health Technology and Informatics, 2025. Verfasst von Inna Skarga-Bandurova und 2 weiteren Autor:innen. Beteiligte Einrichtung: Oxford Brookes University. Hauptthemen laut MEDLINE: Deep Learning, Uterine Cervical Neoplasms, Artificial Intelligence, Image Interpretation, Computer-Assisted. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The study underscores the importance of addressing biases in medical AI models to improve fairness, generalizability, and clinical utility. In this paper, we present a novel framework that combines Explainable AI (XAI) with image darkness assessment to detect and mitigate bias in cervical histology image classification. Four deep learning architectures were employed—AlexNet, ResNet-50, EfficientNet-B0, and DenseNet-121—with EfficientNet-B0 demonstrating the highest accuracy post-mitigation. Grad-CAM and saliency maps were used to identify biases in the models’ predictions. After applying brightness normalisation and synthetic data augmentation, the models shifted focus toward clinically relevant features, improving both accuracy and fairness. Statistical analysis using ANOVA confirmed a reduction in the influence of image darkness on model predictions after mitigation, as evidenced by a decrease in the F-statistic from 120.79 to 14.05, indicating improved alignment of the models with clinically relevant features.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningUterine Cervical NeoplasmsArtificial IntelligenceImage Interpretation, Computer-AssistedHumansFemaleBias

Stichwörter der Autor:innen

Explainable AI (XAI)bias detectionhistopathologyimage darkness

Autor:innen und Institutionen

  • Oxford Brookes University.
  • G.E. Pukhov Institute for Modelling in Energy Engineering.

Zitieren

APA

Skarga-Bandurova, I., Sharifnia, G., & Biloborodova, T. (2025). Bias Detection in Histology Images Using Explainable AI and Image Darkness Assessment. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti250302

Vancouver

Skarga-Bandurova I, Sharifnia G, Biloborodova T. Bias Detection in Histology Images Using Explainable AI and Image Darkness Assessment. Studies in Health Technology and Informatics. 2025. doi:10.3233/shti250302

BibTeX
@article{SkargaBandurova2025,
  author = {Skarga-Bandurova, Inna and Sharifnia, Golshid and Biloborodova, Tetiana},
  title = {Bias Detection in Histology Images Using Explainable AI and Image Darkness Assessment},
  journal = {Studies in Health Technology and Informatics},
  year = {2025},
  doi = {10.3233/shti250302}
}
Kostenlos starten →