A CT Lesion Foundation Model for Non‐Invasive Assessment of Hepatocellular Carcinoma Aggressiveness and Prediction of Recurrence Risk
- 1Department of Radiology & Faculty of Medical Imaging Technology Ruijin Hospital College of Health Science and Technology Shanghai Jiao Tong University School of Medicine Shanghai China
- 2Department of Radiology Xiangya Hospital Central South University Changsha China
Abstract
Hepatocellular carcinoma (HCC) remains a leading cause of cancer‐related mortality worldwide, with high recurrence rates following curative resection. Accurate non‐invasive assessment of tumor aggressiveness and recurrence risk is critical for surgical decision‐making and postoperative surveillance, yet remains challenging with conventional imaging modalities. Here, we developed a CT lesion foundation model (LFM), trained via self‐supervised learning on 32,120 generic lesion and 6,387 HCC‐specific CT patches to capture transferable intra‐lesional imaging phenotypes. Across one internal and two external validation cohorts, the sequentially pretrained LFM (LFM‐Seq) achieved area under the curve values of 0.916, 0.846, and 0.920 for microvascular invasion (MVI) assessment, and 0.833, 0.885 and 0.795 for pathological grade assessment. The model remained robust under scenarios of limited training data and tumor localization perturbations. Furthermore, LFM‐Seq‐derived risk scores stratified recurrence‐free survival (RFS), yielding an externally validated C‐index of 0.701, and remained independently associated with RFS after adjusting for clinicopathological factors. Exploratory biological interpretability analyses showed that LFM‐Seq‐derived risk phenotypes were associated with suppressed lymphocyte‐related immune programs. These results suggest that CT lesion foundation models can support non‐invasive and robust characterization of HCC aggressiveness, recurrence risk stratification and discovery of interpretable prognostic imaging biomarkers.
Related
- Preoperative 3D morphological phenotyping of hepatocellular carcinoma via a foundation model: a non-invasive biomarker for recurrence risk stratification.
- Foundation Model-Enabled Pathology-Informed MRI Model for Preoperative Prediction of Vessels Encapsulating Tumor Clusters in Resectable Hepatocellular Carcinoma: A Multicenter Retrospective Study.
- MIFF-CLIP-HCC: Contrastive Language-Image Pretraining-enhanced multimodal information fusion framework for early recurrence prediction in postoperative hepatocellular carcinoma.
- Preoperative Zinc Deficiency as a Novel Prognostic Predictor for Postoperative Recurrence in Hepatocellular Carcinoma
- Peri-interventional neutrophil extracellular traps and immune surveillance in hepatocellular carcinoma: mechanisms, biomarkers, and recurrence after liver resection or thermal ablation
BibTeX
@article{xia2026lesion,
title = {{A CT Lesion Foundation Model for Non‐Invasive Assessment of Hepatocellular Carcinoma Aggressiveness and Prediction of Recurrence Risk}},
author = {Wei Xia and Zhehan Shen and Mengsi Li and Jianxin Lu and Ji Xia and Yang Cheng and Wenzheng Li and Shunjie Dong and Lin Deng and Ruokun Li},
journal = {Advanced Science},
year = {2026},
month = nov,
doi = {10.1002/advs.78587},
url = {https://doi.org/10.1002/advs.78587},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - A CT Lesion Foundation Model for Non‐Invasive Assessment of Hepatocellular Carcinoma Aggressiveness and Prediction of Recurrence Risk
AU - Wei Xia
AU - Zhehan Shen
AU - Mengsi Li
AU - Jianxin Lu
AU - Ji Xia
AU - Yang Cheng
AU - Wenzheng Li
AU - Shunjie Dong
AU - Lin Deng
AU - Ruokun Li
PY - 2026
DA - 2026/11/01
JO - Advanced Science
DO - 10.1002/advs.78587
UR - https://doi.org/10.1002/advs.78587
AB - Hepatocellular carcinoma (HCC) remains a leading cause of cancer‐related mortality worldwide, with high recurrence rates following curative resection. Accurate non‐invasive assessment of tumor aggressiveness and recurrence risk is critical for surgical decision‐making and postoperative surveillance, yet remains challenging with conventional imaging modalities. Here, we developed a CT lesion foundation model (LFM), trained via self‐supervised learning on 32,120 generic lesion and 6,387 HCC‐specific CT patches to capture transferable intra‐lesional imaging phenotypes. Across one internal and two external validation cohorts, the sequentially pretrained LFM (LFM‐Seq) achieved area under the curve values of 0.916, 0.846, and 0.920 for microvascular invasion (MVI) assessment, and 0.833, 0.885 and 0.795 for pathological grade assessment. The model remained robust under scenarios of limited training data and tumor localization perturbations. Furthermore, LFM‐Seq‐derived risk scores stratified recurrence‐free survival (RFS), yielding an externally validated C‐index of 0.701, and remained independently associated with RFS after adjusting for clinicopathological factors. Exploratory biological interpretability analyses showed that LFM‐Seq‐derived risk phenotypes were associated with suppressed lymphocyte‐related immune programs. These results suggest that CT lesion foundation models can support non‐invasive and robust characterization of HCC aggressiveness, recurrence risk stratification and discovery of interpretable prognostic imaging biomarkers.
SN - 2198-3844
ER -