FMCerNet: frequency-aware multi-label learning for fine-grained cervical cytology screening
4 affiliations
- 1College of Computer Science, Zhejiang University , Hangzhou, Zhejiang 310058,
- 2Department of Pathology, The First Affiliated Hospital, Zhejiang University School of Medicine , Hangzhou, Zhejiang 310003,
- 3School of Software Technology, Zhejiang University , Ningbo, Zhejiang 315100,
- 4International Peace Maternal and Child Health Hospital, School of Medicine, Shanghai Jiao Tong University , Shanghai 200030,
Abstract
Motivation: Automated cytology screening is essential for early cervical cancer prevention, where gigapixel whole-slide images (WSIs) are typically partitioned into tiles for tile-level feature extraction and subsequent WSI-level diagnosis aggregation. However, existing tile-level methods remain constrained by limited input sizes, which frequently obscures sparse and morphologically subtle abnormal cells. Furthermore, these methods commonly formulate tile-level representation learning as a coarse single-label classification task, thereby failing to capture the complex co-occurrence patterns and morphological heterogeneity of cervical cytological abnormalities.
Results: We propose the Frequency-Aware Multi-Label Learning Network (FMCerNet), a tile-level feature extractor equipped with a Large-Field Frequency-Enhanced ViT backbone and a Coarse-to-Fine multi-label classification head. The former incorporates a Dual-Tree Complex Wavelet Transform module to efficiently extract frequency-domain features from large-scale tiles, whereas the latter enhances fine-grained tile representation decoding through hierarchical supervision. Experiments on both public and private datasets demonstrate that FMCerNet improves the accuracy of positive screening and fine-grained subtype classification while maintaining high inference speed.
Availability and implementation: Code at https://github.com/imAzhou/FMCerNet and archived on Zenodo (doi: 10.5281/zenodo.22886105).
Related
- Ultrasound-guided fine-needle aspiration cytology combined with washout fluid thyroglobulin measurement in the diagnosis of cervical lymph node metastasis in papillary thyroid carcinoma: research advances
- Cervical leukoplakia in Ghana: Unveiling diagnostic challenges and informing low-resource screening strategies
- Untimely follow-up after cervical cancer screening: A qualitative study among healthcare providers in Flanders (Belgium)
- Assessing the Capacity of Health Facilities to Deliver Concomitant HPV Vaccination and HPV-Based Cervical Screening Services to Vulnerable Populations – An Implementation Research Protocol
- EEG-based deep learning estimates preterm infants’ cerebral age: Contribution of slow-wave and cross-frequency coupling
BibTeX
@article{zhou2026fmcernet,
title = {{FMCerNet: frequency-aware multi-label learning for fine-grained cervical cytology screening}},
author = {Linyun Zhou and Qiuyu Cai and Xiaoling Wang and Weihan Li and Jian Yang and Mingli Song and Jing Zhang and Feng Zhou and Qihan You and Yudong Wang and Zunlei Feng and Xiuming Zhang},
journal = {Bioinformatics},
year = {2026},
month = oct,
doi = {10.1093/bioinformatics/btag759},
url = {https://academic.oup.com/bioinformatics/article/doi/10.1093/bioinformatics/btag759/8943329},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - FMCerNet: frequency-aware multi-label learning for fine-grained cervical cytology screening
AU - Linyun Zhou
AU - Qiuyu Cai
AU - Xiaoling Wang
AU - Weihan Li
AU - Jian Yang
AU - Mingli Song
AU - Jing Zhang
AU - Feng Zhou
AU - Qihan You
AU - Yudong Wang
AU - Zunlei Feng
AU - Xiuming Zhang
PY - 2026
DA - 2026/10/19
JO - Bioinformatics
DO - 10.1093/bioinformatics/btag759
UR - https://academic.oup.com/bioinformatics/article/doi/10.1093/bioinformatics/btag759/8943329
AB - Motivation: Automated cytology screening is essential for early cervical cancer prevention, where gigapixel whole-slide images (WSIs) are typically partitioned into tiles for tile-level feature extraction and subsequent WSI-level diagnosis aggregation. However, existing tile-level methods remain constrained by limited input sizes, which frequently obscures sparse and morphologically subtle abnormal cells. Furthermore, these methods commonly formulate tile-level representation learning as a coarse single-label classification task, thereby failing to capture the complex co-occurrence patterns and morphological heterogeneity of cervical cytological abnormalities. Results: We propose the Frequency-Aware Multi-Label Learning Network (FMCerNet), a tile-level feature extractor equipped with a Large-Field Frequency-Enhanced ViT backbone and a Coarse-to-Fine multi-label classification head. The former incorporates a Dual-Tree Complex Wavelet Transform module to efficiently extract frequency-domain features from large-scale tiles, whereas the latter enhances fine-grained tile representation decoding through hierarchical supervision. Experiments on both public and private datasets demonstrate that FMCerNet improves the accuracy of positive screening and fine-grained subtype classification while maintaining high inference speed. Availability and implementation: Code at https://github.com/imAzhou/FMCerNet and archived on Zenodo (doi: 10.5281/zenodo.22886105).
SN - 1367-4803
SN - 1367-4811
ER -