EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Breaking Barriers in Thyroid Cytopathology: Harnessing Deep Learning for Accurate Diagnosis

Seo Young Oh iD, Yong Moon Lee iD, Dong Joo Kang, Hyeong Ju Kwon, Sabyasachi Chakraborty iD, Jae Hyun Park iD

Bioengineering · 14.03.2025 · 12 (3) · 293

DOI: 10.3390/bioengineering12030293 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Bioengineering, online erschienen am 14.03.2025. Verfasst von Seo Young Oh und 5 weiteren Autor:innen. Beteiligte Einrichtung: Terenz Co., Ltd., Busan 48060, Republic of Korea. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

We address the application of artificial intelligence (AI) techniques in thyroid cytopathology, specifically for diagnosing papillary thyroid carcinoma (PTC), the most common type of thyroid cancer.

Methods

Our research introduces deep learning frameworks that analyze cytological images from fine-needle aspiration cytology (FNAC), a key preoperative diagnostic method for PTC. The first framework is a patch-level classifier referred as “TCS-CNN”, based on a convolutional neural network (CNN) architecture, to predict thyroid cancer based on the Bethesda System (TBS) category. The second framework is an attention-based deep multiple instance learning (AD-MIL) model, which employs a feature extractor using TCS-CNN and an attention mechanism to aggregate features from smaller-patch-level regions into predictions for larger-patch-level regions, referred to as bag-level predictions in this context.

Results

The proposed TCS-CNN framework achieves an accuracy of 97% and a recall of 96% for small-patch-level classification, accurately capturing local malignancy information. Additionally, the AD-MIL framework also achieves approximately 96% accuracy and recall, demonstrating that this framework can maintain comparable performance while expanding the diagnostic coverage to larger regions through patch aggregation.

Conclusions

This study provides a feasibility analysis for thyroid cytopathology classification and visual interpretability for AI diagnosis, suggesting potential improvements in patient outcomes and reductions in healthcare costs.

Schlagwörter

Stichwörter der Autor:innen

convolutional neural networkdeep learningmultiple instance learningpapillary thyroid cancer

Autor:innen und Institutionen

  • Terenz Co., Ltd., Busan 48060, Republic of Korea.
  • Department of Pathology, College of Medicine, Dankook University, Cheonan 31116, Republic of Korea.
  • Department of Pathology, Wonju Severance Christian Hospital, Wonju College of Medicine, Yonsei University, Seoul 03722, Republic of Korea.
  • Department of Surgery, Wonju Severance Christian Hospital, Wonju College of Medicine, Yonsei University, Wonju 26492, Republic of Korea.

ORCID-Profile: Seo Young Oh, Yong Moon Lee, Sabyasachi Chakraborty, Jae Hyun Park

Zitieren

APA

Oh, S. Y., Lee, Y. M., Kang, D. J., Kwon, H. J., Chakraborty, S., & Park, J. H. (2025). Breaking Barriers in Thyroid Cytopathology: Harnessing Deep Learning for Accurate Diagnosis. Bioengineering, 12(3), 293. https://doi.org/10.3390/bioengineering12030293

Vancouver

Oh SY, Lee YM, Kang DJ, Kwon HJ, Chakraborty S, Park JH. Breaking Barriers in Thyroid Cytopathology: Harnessing Deep Learning for Accurate Diagnosis. Bioengineering. 2025;12(3):293. doi:10.3390/bioengineering12030293

BibTeX
@article{Oh2025,
  author = {Oh, Seo Young and Lee, Yong Moon and Kang, Dong Joo and Kwon, Hyeong Ju and Chakraborty, Sabyasachi and Park, Jae Hyun},
  title = {Breaking Barriers in Thyroid Cytopathology: Harnessing Deep Learning for Accurate Diagnosis},
  journal = {Bioengineering},
  year = {2025},
  volume = {12},
  number = {3},
  pages = {293},
  doi = {10.3390/bioengineering12030293}
}
Kostenlos starten →