Skip to content
DraconiticResearch
Search Sources API
226,742 studies31 sourcesUpdated 36 min ago
Peer-reviewedCC BY-NC-ND 4.0Diagnostic Pathology

Deep learning-powered virtual eosin-based elastic fiber framework imaging improves objective grading and prognostic stratification of lung adenocarcinoma

5 affiliations
  1. 1Department of Pathology, Chongqing Traditional Chinese Medicine Hospital, Chongqing, 400021 China
  2. 2Department of Respiratory and Critical Care Medicine, Chongqing University Jiangjin Hospital, Chongqing, 402260 China
  3. 3Department of Pathology, Chongqing Jiulongpo People’s Hospital, Chongqing, 400050 China
  4. 4Department of Pathology, Dianjiang Hospital of Traditional Chinese Medicine, Chongqing, 408300 China
  5. 5Department of Pathology From College of Basic Medicine, and Molecular Medicine Diagnostic & Testing Center, and Department of Clinical Pathology Laboratory of Pathology Diagnostic Center, Chongqing Medical University, Chongqing, 400016 China

Abstract

Background: Histological grading of lung adenocarcinoma is limited by interobserver variability, particularly when distinguishing noninvasive lepidic growth from invasive patterns. This distinction depends on evaluation of the alveolar elastic fiber framework, which is often insufficiently visualized on routine hematoxylin and eosin (H&E) slides.

Methods: We developed a deep learning framework to improve grading objectivity by computationally visualizing elastic fiber–related architecture from standard H&E images. A dual-stream EfficientNet-B0 classifier was trained using spatially coregistered H&E and eosin-based elastin fluorescence (EBEF) images. To enable clinical deployment without additional staining or fluorescence microscopy, a Pix2Pix-HD generative adversarial network was trained to synthesize high-fidelity virtual EBEF images from H&E alone. The resulting automated pipeline classified histological patterns, quantified pattern proportions, and assigned tumor grades. Performance was evaluated on independent multicenter cohorts, with ablation analyses assessing key pipeline components. Prognostic relevance was explored using a curated TCGA lung adenocarcinoma cohort.

Results: The benchmark dual-stream model using real H&E and EBEF achieved 89.3% tile-level classification accuracy. Using virtual EBEF images, the automated grading pipeline achieved 94.2% accuracy on an external validation cohort, outperforming an H&E-only model (79.6%). Ablation studies confirmed that performance gains depended on informative multimodal input, accurate spatial registration, and attentive feature fusion. AI-assigned grades demonstrated a consistent trend toward association with overall survival in TCGA, with significant stratification observed when low-grade tumors were combined (log-rank p = 0.048).

Conclusions: This framework enables objective grading of lung adenocarcinoma from a single routine H&E slide by enhancing visualization of elastic fiber–related architecture. The approach improves grading consistency and prognostic stratification without altering standard pathology workflows.

Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s13000-026-01837-0.

Related

  1. Prognostic Modeling of Lysosomal Cell Death Signatures in Lung Adenocarcinoma: A Machine Learning Framework for Survival Outcome Prediction
    Runzhe Wang, Qingtao Zhao, Shicheng Liu +5
    International Journal of GenomicsPeer-reviewedCC BY 4.0Full textOncologyPMC
  2. Development and Validation of an Automated Image-Based Method for Prognostic Stratification in Lung Adenocarcinoma using Immunohistochemical Markers
    F. Carraro, M. Sanguesa, A. Irisarri +7
    medRxivPreprintCC BYFull textEpidemiology & public healthmedRxiv
  3. An entosis-informed prognostic index for survival stratification and molecular characterization in lung adenocarcinoma.
    Yinqin Liu, Wen Zhao, Shuyi Mu +10
    Frontiers in cell and developmental biologyPeer-reviewedCC BYFull textGeneral medicineEurope PMC
  4. Leveraging deep learning for prognostic stratification of hepatocellular carcinoma using histopathological images
    Dengbin Wu, Jiaoyue Zhang, Mian Shi +4
    Frontiers in oncologyPeer-reviewedCC BYFull textOncologyEurope PMC
  5. Radiomics-based machine learning models decipher heterogeneity of anoikis-related molecular subtypes for prognostic stratification in head and neck squamous cell carcinoma
    Linyong Wu, Dayou Wei, Wubiao Chen +4
    Translational cancer researchPeer-reviewedCC BY-NC-NDFull textGeneral medicineEurope PMC
BibTeX
@article{wang2026deep,
  title = {{Deep learning-powered virtual eosin-based elastic fiber framework imaging improves objective grading and prognostic stratification of lung adenocarcinoma}},
  author = {Cheng-Long Wang and Shi-Peng Lei and Yan Lu and Ling-Feng Zou and Jian-Bo Xiong and Jing-Wen Li and Xin Ouyang and Lu He and Yi-Ying Luo and Li Zhang and Xiao-Jing Cao and Shan-Shan Yu},
  journal = {Diagnostic Pathology},
  year = {2026},
  month = aug,
  doi = {10.1186/s13000-026-01837-0},
  url = {https://doi.org/10.1186/s13000-026-01837-0},
  note = {CC BY-NC-ND 4.0},
}
RIS
TY  - JOUR
TI  - Deep learning-powered virtual eosin-based elastic fiber framework imaging improves objective grading and prognostic stratification of lung adenocarcinoma
AU  - Cheng-Long Wang
AU  - Shi-Peng Lei
AU  - Yan Lu
AU  - Ling-Feng Zou
AU  - Jian-Bo Xiong
AU  - Jing-Wen Li
AU  - Xin Ouyang
AU  - Lu He
AU  - Yi-Ying Luo
AU  - Li Zhang
AU  - Xiao-Jing Cao
AU  - Shan-Shan Yu
PY  - 2026
DA  - 2026/08/17
JO  - Diagnostic Pathology
DO  - 10.1186/s13000-026-01837-0
UR  - https://doi.org/10.1186/s13000-026-01837-0
AB  - Background: Histological grading of lung adenocarcinoma is limited by interobserver variability, particularly when distinguishing noninvasive lepidic growth from invasive patterns. This distinction depends on evaluation of the alveolar elastic fiber framework, which is often insufficiently visualized on routine hematoxylin and eosin (H&E) slides.  Methods: We developed a deep learning framework to improve grading objectivity by computationally visualizing elastic fiber–related architecture from standard H&E images. A dual-stream EfficientNet-B0 classifier was trained using spatially coregistered H&E and eosin-based elastin fluorescence (EBEF) images. To enable clinical deployment without additional staining or fluorescence microscopy, a Pix2Pix-HD generative adversarial network was trained to synthesize high-fidelity virtual EBEF images from H&E alone. The resulting automated pipeline classified histological patterns, quantified pattern proportions, and assigned tumor grades. Performance was evaluated on independent multicenter cohorts, with ablation analyses assessing key pipeline components. Prognostic relevance was explored using a curated TCGA lung adenocarcinoma cohort.  Results: The benchmark dual-stream model using real H&E and EBEF achieved 89.3% tile-level classification accuracy. Using virtual EBEF images, the automated grading pipeline achieved 94.2% accuracy on an external validation cohort, outperforming an H&E-only model (79.6%). Ablation studies confirmed that performance gains depended on informative multimodal input, accurate spatial registration, and attentive feature fusion. AI-assigned grades demonstrated a consistent trend toward association with overall survival in TCGA, with significant stratification observed when low-grade tumors were combined (log-rank p = 0.048).  Conclusions: This framework enables objective grading of lung adenocarcinoma from a single routine H&E slide by enhancing visualization of elastic fiber–related architecture. The approach improves grading consistency and prognostic stratification without altering standard pathology workflows.  Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s13000-026-01837-0.
SN  - 1746-1596
ER  -