Related Experiment Video
Updated: Jul 16, 2026
Using Computer-based Image Analysis to Improve Quantification of Lung Metastasis in the 4T1 Breast Cancer Model
Published on: October 2, 2020
Development and application of a prognostic model based on radiomics and artificial intelligence for patients with
Congying Zheng1,2, Xinyuan Yang2,3,4, Musen Ye5
1Department of Oncology, Shenzhen Key Laboratory of Gastrointestinal Cancer Translational Research, Cancer Institute, Peking University Shenzhen Hospital, Shenzhen-Peking University-Hong Kong University of Science and Technology Medical Center, Shenzhen, China.
Introduction:
Lung cancer with brain metastasis (LCBM) impairs survival in lung adenocarcinoma. High postoperative recurrence rates highlight the necessity of accurate prognostic tools. This study aimed to develop an integrated radiomics-clinical model to improve survival prediction in lung adenocarcinoma patients with brain metastasis.
Methods:
The cohort of 176 patients with LCBM was randomly divided into a training set (n=123) and a test set (n=53). The identification of clinical risk factors was performed using both univariate and multivariate logistic regression analyses. A radiomics model was developed based on radiomic features extracted from preoperative magnetic resonance imaging (MRI), following selection with Least Absolute Shrinkage and Selection Operator (LASSO) regression. The performance of the combined nomogram, which integrated significant clinical and radiomic features, was evaluated by the area under the receiver operating characteristic curve (AUC), along with calibration and decision curve analyses.
Results:
Multivariate analysis established the EGFR mutation status, number of brain metastases, and Lung-molGPA score as independent prognostic determinants. Performance evaluation of the radiomics model yielded AUCs of 0.862 in the training set and 0.829 in the test set, indicating robust diagnostic performance. The combined nomogram demonstrated superior predictive performance, with AUC values of 0.904 and 0.874 in the training and test sets, respectively, along with good calibration and clinical utility in both cohorts.
Discussion:
These findings demonstrate the combined utility of integrating radiomics with clinical parameters to enhance prognostic accuracy, enabling personalized treatment stratification in LCBM and improving clinical decision-making and risk stratification.
