A Robust Framework for Domain-Generalized Classification of Ovarian Cancer Histology Images

<b>Background:</b> In computational pathology (CP) analysis, computational efficiency and precise classification outcomes are paramount for robust and scalable solutions. Despite recent advancements in deep-learning frameworks for Whole-Slide Images (WSIs), the heterogeneity of WSIs across different domains poses considerable challenges for developing models with robust generalization capabilities. This study presents the WSI-P2P (Whole-Slide Imaging-Patch to Prediction), aimed at addressing these challenges. <b>Methods:</b> WSI-P2P leverages downscaled patch sampling and Multiple-Instance Learning (MIL) with transfer learning to optimize resource usage while maintaining a competitive performance. Within WSI-P2P, we introduce the <i>K-TOP</i> MIL aggregator, a variant of the MIL attention-based aggregator, which selectively processes the most informative <i>K</i> instances. The framework features an online, adaptive feature extractor that fine-tunes pre-trained models in an end-to-end manner, addressing multi-centered dataset variability. <b>Results:</b> WSI-P2P achieves state-of-the-art accuracy, demonstrating superior domain adaptability and computational efficiency. WSI-P2P is validated by employing several dataset splits and down-sampled patch variations, illustrating its potential as a scalable and reliable tool in clinical settings and large-scale histological studies. The framework achieved a maximum score of 95.89% AUROC and a test accuracy of 77.67% without attention, further improving to approximately 100% AUROC and a test accuracy of 95.72% recorded with the <i>K-TOP</i> MIL aggregator. Further, for intra-domain generalization experiments, WSI-P2P recorded a consistent performance across domains, validating its domain generalization capabilities. The <i>K-TOP</i> MIL aggregator also demonstrated 2.3× computational efficiency as compared to base aggregators. <b>Conclusions:</b> The proposed framework outperforms traditional offline feature extraction methods, ensuring high discriminative ability even when exposed to data from diverse distributions. WSI-P2P demonstrates excellent performance between subtype classifications, positioning it as a reliable tool for large-scale histological studies.