Sep 2026· Medical and Biological Engineering and Computing· 0 citations· 27 references
Medicine
TL;DR
Neighbor-Constrained MIL (NCMIL), a dual-path MIL framework that combines a Nyströmformer global encoder with Neighbor-Constrained Attention (NCA) and introduces three design choices aimed at spatially structured WSI evidence, achieves the best overall performance among the evaluated baselines.
SlideMix is proposed, a model-agnostic multimodal augmentation framework for MIL-based WSI analysis that improves accuracy and generalization in most settings and compares favorably with established augmentation baselines, providing a simple plug-and-play approach for more robust and scalable digital pathology models.
Chad Wong, Si-Cheng Chen, Tian-Yi Zhang et al.· 0 citations
CRAFT (Coarse-to-fine Region-Adaptive Feature Tokenization), a DINO-based framework that learns image-dependent mixed-scale representations by using self-supervised attention to selectively refine informative regions while preserving coarse context, together with a symmetric cross-scale regularization objective that en...
Anabel Stammer, Valay Bundele, Mehran Hosseinzadeh et al.· 0 citations
Whole slide image (WSI) analysis in computational pathology follows a multiple instance learning (MIL) pipeline where patch embeddings are extracted independently and aggregated for slide-level prediction, but within-slide variance from staining, scanner, and local texture can overwhelm the discriminative signal. We pr...
Hao-Yu He, B. Tessier-Cloutier, Yang Wang et al.· 0 citations
Automating the analysis of whole-slide images (WSIs), a key step in cancer diagnosis, has high clinical value, as it can reduce pathologist's workload while improving diagnosis accuracy. Recently, vision-language models have shown promising performance for patch-level classification without requiring any annotation, ye...
T. Godelaine, M. Dausort, Karim El Khoury et al.· 0 citations
In weakly supervised Whole Slide Image (WSI) classification, feature extractors typically compress each image tile into a single global embedding. Consequently, slide-level aggregators are restricted to this coarse tile scale, concealing fine-grained sub-tile evidence from the attention mechanism. We introduce DI-MIL,...
Run-Sheng Liu, Cheng Jin, Hao Jiang et al.· 0 citations
Automated triage of whole-slide cervical cytology images remains constrained by the heavy dependence on single-cell bounding boxes or pixel-level annotations, which introduce significant bottlenecks, high expert overhead, and observer variability. To address these challenges, we present a weakly-supervised, detection-f...
S. Mayala, S. Sang'udi, D. Mzurikwao et al.· medRxiv· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
