An interpretable multi‐task whole‐slide histopathology AI model for non‐small cell lung cancer: Cross‐cohort generalisation, spatial attention–transcriptomic integration, and molecular–immune profiling
Clinical and Translational Medicine · 2026-07 · 16 (7)
DOI: 10.1002/ctm2.70744 ↗Kurz gefasst
Fachartikel in der Zeitschrift Clinical and Translational Medicine, online erschienen am 23.07.2026. Verfasst von Renyi Lu und 12 weiteren Autor:innen. Beteiligte Einrichtung: Department of Oncology, Zhujiang Hospital, The First School of Clinical Medicine, Southern Medical University; Donghai County People's Hospital (Affi…. Hauptthemen laut MEDLINE: Carcinoma, Non-Small-Cell Lung, Lung Neoplasms. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 96 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Research Start-up Fund at Zhujiang Hospital.
Abstract (Originalsprache)
Background
Tumour‐node‐metastasis staging does not fully explain prognostic heterogeneity in non‐small cell lung cancer. We evaluated whether haematoxylin‐and‐eosin whole‐slide images could estimate histological subtype, pathological stage probabilities, survival risk and spatially grounded biological associations.
Methods
SparseAGE‐MTL, a weakly supervised multi‐task multiple‐instance learning model with a shared projection‐topology encoder and endpoint‐specific heads, was trained and benchmarked in 954 The Cancer Genome Atlas cases using seven pathology feature spaces and 18 comparator models. External evaluation used 948 tissue‐microarray and 324 whole‐slide cases. Attention maps were co‐registered with 10x Visium spatial transcriptomics and integrated with bulk transcriptomics, immune‐infiltration estimates and ESTIMATE scores. Analyses included paired model comparisons, false‐discovery‐rate correction, Cox models, calibration assessment and decision curve analysis.
Results
In the CONCH feature space, SparseAGE‐MTL achieved 93.73% accuracy, 98.19% area under the receiver‐operating‐characteristic curve and 93.08% F1‐score for adenocarcinoma/squamous cell carcinoma classification in internal benchmarking; external area‐under‐the‐curve values were approximately .91 and .82. Stage estimation had lower discrimination, with external overall area under the curve approximately .70 and cohort‐dependent calibration. Risk‐score‐defined groups differed in overall survival in both histological subtypes and showed similar external trends. High‐attention regions were enriched at tumour–stroma or tumour–immune interfaces and were associated with B‐cell, fibroblast, C1QC, COL1A1, epithelial–mesenchymal transition, metastasis and hypoxia signals. Higher risk cases showed malignant pathway activation, lower immune/stromal scores, higher tumour purity and subtype‐specific immune/stromal differences. Adding the risk score to the clinical model increased external pooled concordance index from approximately .620 to .672.
Conclusions
In retrospective cohorts, SparseAGE‐MTL generated subtype‐classification, stage‐probability and survival‐risk outputs from routine pathology images. Subtype classification had higher numerical performance than stage estimation. Survival‐risk and attention outputs were associated with outcome and spatial/transcriptomic features, but prospective, treatment‐annotated validation is required before clinical use. Key points SparseAGE‐MTL is a weakly supervised multi‐task MIL framework that jointly performs NSCLC subtype classification, stage prediction, and survival risk estimation from routine H&E slides using only slide-level labels. The model achieves stable competitive performance across seven feature spaces and 18 comparators, with external validation demonstrating robust generalization to independent WSI and TMA cohorts. Attention hotspots co‐localize with tumor‐stroma/immune interfaces and spatial transcriptomic signatures of EMT, hypoxia, and C1QC/COL1A1 enrichment, providing biologically grounded interpretability. High‐risk groups exhibit activated malignant pathways, lower immune/stromal scores, higher tumor purity, and incremental prognostic value beyond clinical variables.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Oncology, Zhujiang Hospital, The First School of Clinical Medicine, Southern Medical University; Donghai County People's Hospital (Affiliated Kangda College of Nanjing Medical University), Lianyungang, Chi…
- Department of Urology, Changhai Hospital, Naval Medical University (Second Military Medical University), Shanghai, China.
- School of Stomatology, Southern Medical University, Guangzhou, Guangdong, China.
- Department of Oncology, Sichuan Provincial People's Hospital, University of Electronic Science and Technology of China, Chengdu, Sichuan, China.
- Department of Radiology, Sichuan Provincial People's Hospital, University of Electronic Science and Technology of China, Chengdu, Sichuan, China.
- Department of Oncology, Zhujiang Hospital, Southern Medical University, Guangzhou, Guangdong, China.
ORCID-Profile: Renyi Lu, Anqi Lin, Aimin Jiang, Yifeng Bai, Jian Zhang, Quan Cheng, Qinglin Li, Peng Luo
Förderung
- Research Start-up Fund at Zhujiang Hospital
Zitieren
APA
Lu, R., Lin, A., Jiang, A., Feng, Y., Fang, X., Shen, J., Bai, Y., Peng, S., Zhang, J., Cheng, Q., Feng, S., Li, Q., & Luo, P. (2026). An interpretable multi‐task whole‐slide histopathology AI model for non‐small cell lung cancer: Cross‐cohort generalisation, spatial attention–transcriptomic integration, and molecular–immune profiling. Clinical and Translational Medicine, 16(7). https://doi.org/10.1002/ctm2.70744
Vancouver
Lu R, Lin A, Jiang A, Feng Y, Fang X, Shen J, et al. An interpretable multi‐task whole‐slide histopathology AI model for non‐small cell lung cancer: Cross‐cohort generalisation, spatial attention–transcriptomic integration, and molecular–immune profiling. Clinical and Translational Medicine. 2026;16(7). doi:10.1002/ctm2.70744
BibTeX
@article{Lu2026,
author = {Lu, Renyi and Lin, Anqi and Jiang, Aimin and Feng, Yuying and Fang, Xiuhui and Shen, Junyi and Bai, Yifeng and Peng, Shengkun and Zhang, Jian and Cheng, Quan and Feng, Suyin and Li, Qinglin and Luo, Peng},
title = {An interpretable multi‐task whole‐slide histopathology AI model for non‐small cell lung cancer: Cross‐cohort generalisation, spatial attention–transcriptomic integration, and molecular–immune profiling},
journal = {Clinical and Translational Medicine},
year = {2026},
volume = {16},
number = {7},
doi = {10.1002/ctm2.70744}
}
EUVIMED