Bag-of-Visual-Words for Spatial Mapping of Lung Adenocarcinoma Growth Patterns

2026-08-05

AI Summary

This study addresses the challenge of constructing spatial maps of histological growth patterns in lung adenocarcinoma whole-slide images that align with clinical definitions. The authors propose a weakly supervised bag-of-visual-words (BoVW) pipeline that integrates frozen foundation model embeddings with sparsely annotated regions to derive interpretable pattern prototypes, enabling whole-slide spatial mapping via sliding-window matching. This work represents the first integration of BoVW with foundation model embeddings for modeling lung adenocarcinoma growth patterns, producing clinically aligned maps while preserving tissue heterogeneity. Evaluated on 87 CPTAC-LUAD patient slides, the method achieves a balanced accuracy of 0.974 for tumor versus healthy tissue classification and outperforms supervised baselines in histological grading, demonstrating its efficacy.

Abstract

Spatial mapping of lung adenocarcinoma (LUAD) growth patterns across whole slide images (WSIs) requires resolving architectural context at the region level, yet existing methods operate at the individual tile level and produce generic morphological clusters rather than clinically defined pattern maps. We propose a weakly supervised Bag-of-Visual-Words (BoVW) pipeline that learns a visual vocabulary from frozen foundation model embeddings extracted from a small set of annotated regions of interest (ROIs). Pattern prototypes are constructed as mean BoVW histograms of same-label ROIs and used for nearest-prototype classification of sliding-window regions under Jensen--Shannon divergence. The resulting predictions are projected onto the WSI tile grid to produce interpretable spatial pattern maps. We evaluate the method on 87 CPTAC-LUAD patients using three foundation model encoders and multiple vocabulary sizes on two clinically motivated tasks. For tumour/healthy classification, the best configuration achieves a balanced accuracy of $0.974$ with H-Optimus-1, approaching the $0.987$ obtained by a supervised SVM trained on mean-pooled WSI embeddings. For binary histologic grade classification, the BoVW pipeline achieves higher balanced accuracy than the supervised baseline for all encoders, suggesting that ROI-level pattern decomposition preserves grade-relevant heterogeneity that is attenuated by global mean pooling.

Problem

• lung adenocarcinoma
• spatial mapping
• growth patterns
• whole slide images
• architectural context

Innovation

• Bag-of-Visual-Words
• weakly supervised learning
• spatial mapping
• foundation model embeddings
• lung adenocarcinoma