# Margin-Consistent Deep Subtyping of Invasive Lung Adenocarcinoma via Perturbation Fidelity in Whole-Slide Image Analysis

Source: [Margin-Consistent Deep Subtyping of Invasive Lung Adenocarcinoma via Perturbation Fidelity in Whole-Slide Image Analysis](<https://www.sota2.com/research/paper/69afa57d2d603e6de3e439d6>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-11T05:00:04.181Z

Authors: Meghdad Sabouri Rad, Junze (Vincent) Huang, Mohammad Mehdi Hosseini, Rakesh Choudhary, Saverio J. Carello, Ola El-Zammar, Michel R. Nasr, Bardia Rodd

Published: 2026-02-27T20:47:15.000Z

## Abstract

Whole-slide image classification for invasive lung adenocarcinoma subtyping remains vulnerable to real-world imaging perturbations that undermine model reliability at the decision boundary. We propose a margin consistency framework evaluated on 203,226 patches from 143 whole-slide images spanning five adenocarcinoma subtypes in the BMIRDS-LUAD dataset. By combining attention-weighted patch aggregation with margin-aware training, our approach achieves robust feature-logit space alignment measured by Kendall correlations of 0.88 during training and 0.64 during validation. Contrastive regularization, while effective at improving class separation, tends to over-cluster features and suppress fine-grained morphological variation; to counteract this, we introduce Perturbation Fidelity (PF) scoring, which imposes structured perturbations through Bayesian-optimized parameters. Vision Transformer-Large achieves 95.20 +/- 4.65% accuracy, representing a 40% error reduction from the 92.00 +/- 5.36% baseline, while ResNet101 with an attention mechanism reaches 95.89 +/- 5.37% from 91.73 +/- 9.23%, a 50% error reduction. All five subtypes exceed an area under the receiver operating characteristic curve (AUC) of 0.99. On the WSSS4LUAD external benchmark, ResNet50 with an attention mechanism attains 80.1% accuracy, demonstrating cross-institutional generalizability despite approximately 15-20% domain-shift-related degradation and identifying opportunities for future adaptation research.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Classification on BMIRDS-LUAD](<https://www.sota2.com/research/sota/classification-on-bmirds-luad.md>): Classification; BMIRDS-LUAD; Accuracy: 95.89.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.06650>)

[PDF](<https://arxiv.org/pdf/2603.06650.pdf>)
