A Multiple Instance Learning Framework for Estradiol Level Classification in TCT Whole Slide Images

Objective: Estradiol level assessment is critical for the management of gynecological endocrine disorders. The current gold standard, serum hormone testing, is limited by its invasiveness, variability due to physiological fluctuations, and limited accessibility in primary care settings. This study investigates the feasibility of using Multiple Instance Learning (MIL) algorithms for the non-invasive classification of estradiol levels based on Thinprep Cytologic Test (TCT) whole slide images (WSIs). Methods: This retrospective study included 1,171 patient samples with paired TCT images and serum estradiol data. Samples were categorized into positive (n = 713) and negative (n = 458) groups using a threshold of 40 pg/mL. Each WSI was treated as a “bag,” with extracted image patches constituting “instances.” Under a weakly supervised MIL model, the binary classification performance of four models (AB-MIL, DS-MIL, Trans-MIL, and DTFD-MIL) for estradiol levels was systematically evaluated. Results Among the four MIL models, AB-MIL demonstrated superior performance, achieving a test accuracy of 0.706 ± 0.048, a macro-AUC of 0.747 ± 0.024, and a macro-F1 score of 0.692 ± 0.043. Conclusion This study successfully established an MIL-based model for estradiol level classification using TCT WSIs, confirming the technical feasibility of non-invasively assessing estradiol status through the analysis of cervical cell morphological features. This approach provides a methodological foundation for developing novel auxiliary diagnostic tools in endocrinology that can be integrated with routine cervical cancer screening for concurrent evaluation.