Bi-Regional Machine Learning Radiomics Based on CT Noninvasively Predicts LOX Expression Level and Overall Survival
Kexin Gao1,2, Musa Yaermaimaiti3, Yiwei Wang1
1Department of General Surgery, Shanghai Sixth People's Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai, China.
Objective:
We aimed to assess the association between lysyl oxidases (LOX) expression levels and the prognosis of patients with Hepatocellular carcinoma (HCC) and to establish a CT-based bi-regional radiomics model that can discriminate LOX expression level using The Cancer Imaging Archive (TCIA) and The Cancer Genome Atlas (TCGA) database.
Methods:
294 HCC samples were downloaded from TCGA for gene-based prognostic analysis. Meanwhile, the underlying molecular mechanism of LOX expression and its relationship with the immune microenvironment was investigated. Thirty-four cases that had preoperative computed tomography (CT) images stored in TCIA with genomic data in TCGA were used for radiomics feature extraction and model construction. The association of this combined LOX-based radiomics model with HCC prognosis was evaluated.
Results:
The expression of LOX in tumor tissue significantly correlated with overall survival (OS). LOX was involved in the regulation of immune response and tumor invasion and metastasis. Two radiomic models were developed using the least absolute shrinkage and selection operator (LASSO) regression analysis. The model of the whole-tumor region has a good predicting effect in LOX expression, with the area under the receiver operating characteristic (ROC) curve being 0.775 in the training set, while the average under the curve (AUC) value of the 5-fold cross-validation was 0.754. The model of the whole-tumor and peri-tumor region also has a good predicting effect in LOX expression, with the area under the ROC curve being 0.800 in the training set, while the average AUC value of the 5-fold cross-validation was 0.796. The calibration curves and the Hosmer-Lemeshow test revealed consistency of the prediction probability acquired by our models and the true value of LOX expression, while the decision curve analysis (DCA) curve showed that both models had clinical practicability.
Conclusion:
LOX expression can influence the prognosis of patients with HCC, which can be predicted noninvasively by CT image-based radiomics.
More Related Videos
07:15Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
07:13Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
