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

• URL: http://arxiv.org/abs/2603.06650v1
• Date: Fri, 27 Feb 2026 20:47:15 GMT
• Title: Margin-Consistent Deep Subtyping of Invasive Lung Adenocarcinoma via Perturbation Fidelity in Whole-Slide Image Analysis
• Abstract summary: Whole-slide image classification for invasive lung adenocarcinoma subtyping remains vulnerable to real-world imaging perturbations.<n>We propose a margin consistency framework evaluated on 203,226 patches from 143 whole-slide images.<n>Our approach achieves robust feature-logit space alignment measured by Kendall correlations of 0.88 during training and 0.64 during validation.
• Score: 15.00702842825464
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• 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.

Related papers

• RobustDefect-LLM: Explainable and Robustness-Aware Industrial Surface Defect Classification with Decision Support and AI-Assisted Reporting [0.0]
  The RobustDefect-LLM framework integrates deep-learning classification, operator-facing visual evidence, confidence-aware decision support, and controlled AI-assisted reporting.<n>MobileNetV3-Large achieved the highest numerical test accuracy (99.26%) and macro F1-score (0.9926), with a 95% accuracy CI of 0.9815-1.0000.
  arXiv  Detail & Related papers  (2026-08-09T09:08:13Z)
• Searching for Robust Augmentations to Improve Out-of-Domain Generalization in Dermoscopic Skin Cancer Classification [35.18016233072556]
  We study how data augmentation improves robustness of a malignant binary-non-malignant generalization.<n>Data augmentation was searched on a multi-source ISIC Archive collection.
  arXiv  Detail & Related papers  (2026-07-29T11:06:06Z)
• A Pathology Foundation Model for Gastric Cancer with Real-World Validation [47.19073124722865]
  We present GRACE, a Gastric-specific foundation model for Real-world Assessment and Clinical dEcision support.<n> GRACE was developed from multicenter gastric pathology datasets totaling 48,364 primarily HE-stained whole-slide images from 37,493 patients.<n>When calibrated to maintain non-inferior performance to senior pathologists, the AI-assisted workflow could triage 60.7% of atrophy and 82.7% of intestinal metaplasia cases.
  arXiv  Detail & Related papers  (2026-06-03T12:18:06Z)
• A Clinically Validated Foundation Model for Comprehensive Lung Pathology Interpretation [36.95302007351382]
  PulmoFoundation is a prospectively validated, randomized controlled trial (RCT)-evaluated foundation model for comprehensive lung pathology assessment.<n>Our model achieves clinical-grade performance in core diagnostic tasks across biopsy, frozen section, and surgical resection slides.
  arXiv  Detail & Related papers  (2026-05-25T14:04:56Z)
• Cross-Stage Attention Multi-Expert Network for Radiologist-Inspired Breast Ultrasound Diagnosis [6.727825549527392]
  This paper proposes the Cross-Stage Attention Mixture-of-Experts Network (CSA-MoE-Net) to improve feature representation and classification accuracy.<n>It adopts a Cross-Stage Attention-enhanced ResNet-18 as the backbone, in which the Cross-Stage Attention module adaptively recalibrates multi-level features.<n> Experiments on a balanced dataset of 2,129 breast ultrasound images show that, averaged over 20 independent runs, the model achieves an accuracy of 96.33%, precision of 94.09%, recall of 98.53%, F1-score of 96.25%, and
  arXiv  Detail & Related papers  (2026-05-25T07:20:13Z)
• Error-Decomposed Class-Conditional Fusion for Statistically Guaranteed Hard-Category Robust Perception [3.4563147061792527]
  This paper formally defines this pervasive vulnerability as the Hard-Category Reliability Problem (HCRP)<n>We propose Error-Decomposed Class-Conditional Fusion (ED-CCF), an elegant decision-layer inference framework.
  arXiv  Detail & Related papers  (2026-05-17T18:39:05Z)
• Beyond Localization: A Comprehensive Diagnosis of Perspective-Conditioned Spatial Reasoning in MLLMs from Omnidirectional Images [52.50752250573993]
  Multimodal Large Language Models (MLLMs) show strong visual perception, yet remain limited in reasoning about space under changing viewpoints.<n>We study this challenge as Perspective-Conditioned Spatial Reasoning (PCSR) in 360-degree omnidirectional images.<n>We introduce PCSR-Bench, a diagnostic benchmark of 84,373 question-answer pairs from 2,600 omnidirectional images.
  arXiv  Detail & Related papers  (2026-05-12T17:11:17Z)
• KAYRA: A Microservice Architecture for AI-Assisted Karyotyping with Cloud and On-Premise Deployment [0.0]
  KAYRA is an end-to-end karyotyping system that operates inside the operational constraints of a clinical cytogenetic laboratory.<n>KAYRA is architected as a containerized microservice pipeline whose ML stack combines an EfficientNet-B5 + U-Net semantic segmenter, a Mask R-CNN (ResNet-50 + FPN) instance, and a ResNet-18 instance.<n>A pilot clinical evaluation against two commercial reference karyotyping systems on 459 chromosomes shows strong empirical performance.
  arXiv  Detail & Related papers  (2026-04-29T16:35:31Z)
• LAW & ORDER: Adaptive Spatial Weighting for Medical Diffusion and Segmentation [3.6879658915311997]
  Learnable Adaptive Weighter (LAW) predicts per-pixel loss modulation from features and masks for diffusion training.<n>ORDER applies selective bidirectional skip attention at late decoder stages for efficient segmentation.<n>Experiments on polyp and kidney tumor datasets demonstrate that LAW achieves 20% FID generative improvement.
  arXiv  Detail & Related papers  (2026-03-05T04:20:32Z)
• Distilled HuBERT for Mobile Speech Emotion Recognition: A Cross-Corpus Validation Study [0.0]
  Speech Emotion Recognition (SER) has significant potential for mobile applications.<n>This paper presents a mobile-efficient SER system based on DistilHuBERT.<n>Cross-corpus evaluation on RAVDESS reveals that the theatrical nature of acted emotions causes predictions to cluster by arousal level rather than by specific emotion categories.
  arXiv  Detail & Related papers  (2025-12-29T12:53:39Z)
• Breast Cell Segmentation Under Extreme Data Constraints: Quantum Enhancement Meets Adaptive Loss Stabilization [1.044992176213712]
  Our framework achieves a Dice score of 95.5% +/- 0.3% and an IoU of 91.2% +/- 0.4% using just 599 training images.<n>By achieving groundbreaking performance with limited annotations, our approach significantly reduces the medical expert time required for dataset creation.
  arXiv  Detail & Related papers  (2025-12-02T00:45:21Z)
• Hard Spatial Gating for Precision-Driven Brain Metastasis Segmentation: Addressing the Over-Segmentation Paradox in Deep Attention Networks [0.0]
  We introduce the Spatial Gating Network (SG-Net), a precision-first architecture employing hard spatial gating mechanisms.<n> SG-Net enforces strict feature selection to aggressively suppress background artifacts while preserving tumor features.<n>These findings establish hard spatial gating as a robust solution for precision-driven lesion detection.
  arXiv  Detail & Related papers  (2025-11-27T16:41:27Z)
• DRBD-Mamba for Robust and Efficient Brain Tumor Segmentation with Analytical Insights [54.87947751720332]
  Accurate brain tumor segmentation is significant for clinical diagnosis and treatment.<n>Mamba-based State Space Models have demonstrated promising performance.<n>We propose a dual-resolution bi-directional Mamba that captures multi-scale long-range dependencies with minimal computational overhead.
  arXiv  Detail & Related papers  (2025-10-16T07:31:21Z)
• HANS-Net: Hyperbolic Convolution and Adaptive Temporal Attention for Accurate and Generalizable Liver and Tumor Segmentation in CT Imaging [2.7205074719266062]
  Accurate liver and tumor segmentation on abdominal CT images is critical for reliable diagnosis and treatment planning.<n>We introduce Hyperbolic-convolutions Adaptive-temporal-attention with Neural-representation and Synaptic-plasticity Network (HANS-Net)<n>HANS-Net combines hyperbolic convolutions for hierarchical geometric representation, a wavelet-inspired decomposition module for multi-scale texture learning, and an implicit neural representation branch.
  arXiv  Detail & Related papers  (2025-07-15T13:56:37Z)
• Correcting Class Imbalances with Self-Training for Improved Universal Lesion Detection and Tagging [43.06199185109424]
  Universal lesion detection and tagging (ULDT) in CT studies is critical for tumor burden assessment and tracking the progression of lesion status (growth/shrinkage) over time.<n>Prior work used the DeepLesion dataset (4,427 patients, 10,594 studies, 32,120 CT slices, 32,735 lesions, 8 body part labels) for algorithmic development, but this dataset is not completely annotated and contains class imbalances.<n>We developed a self-training pipeline for ULDT using a limited 11.5% subset of DeepLesion.
  arXiv  Detail & Related papers  (2025-04-07T15:57:03Z)
• Patch-Level Contrasting without Patch Correspondence for Accurate and Dense Contrastive Representation Learning [79.43940012723539]
  ADCLR is a self-supervised learning framework for learning accurate and dense vision representation. Our approach achieves new state-of-the-art performance for contrastive methods.
  arXiv  Detail & Related papers  (2023-06-23T07:38:09Z)
• Corneal endothelium assessment in specular microscopy images with Fuchs' dystrophy via deep regression of signed distance maps [48.498376125522114]
  This paper proposes a UNet-based segmentation approach that requires minimal post-processing. It achieves reliable CE morphometric assessment and guttae identification across all degrees of Fuchs' dystrophy.
  arXiv  Detail & Related papers  (2022-10-13T15:34:20Z)
• (Certified!!) Adversarial Robustness for Free! [116.6052628829344]
  We certify 71% accuracy on ImageNet under adversarial perturbations constrained to be within a 2-norm of 0.5. We obtain these results using only pretrained diffusion models and image classifiers, without requiring any fine tuning or retraining of model parameters.
  arXiv  Detail & Related papers  (2022-06-21T17:27:27Z)
• EMT-NET: Efficient multitask network for computer-aided diagnosis of breast cancer [58.720142291102135]
  We propose an efficient and light-weighted learning architecture to classify and segment breast tumors simultaneously. We incorporate a segmentation task into a tumor classification network, which makes the backbone network learn representations focused on tumor regions. The accuracy, sensitivity, and specificity of tumor classification is 88.6%, 94.1%, and 85.3%, respectively.
  arXiv  Detail & Related papers  (2022-01-13T05:24:40Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.