CT-based deep learning model: a novel approach to the preoperative staging in patients with peritoneal metastasis.

Peritoneal metastasis (PM) is a frequent manifestation of advanced abdominal malignancies. Accurately assessing the extent of PM before surgery is essential for patients to receive optimal treatment. Therefore, we propose to construct a deep learning (DL) model based on enhanced computed tomography (CT) images to stage PM preoperatively in patients. All 168 patients with PM underwent contrast-enhanced abdominal CT before either open surgery or laparoscopic exploration, and peritoneal cancer index (PCI) was used to evaluate patients during the surgical procedure. DL features were extracted from portal venous-phase abdominal CT scans and subjected to feature selection using the Spearman correlation coefficient and LASSO. The performance of models for preoperative staging was assessed in the validation cohort and compared against models based on clinical and radiomics (Rad) signature. The DenseNet121-SVM model demonstrated strong patient discrimination in both the training and validation cohorts, achieving AUC was 0.996 in training and 0.951 validation cohort, which were both higher than those of the Clinic model and Rad model. Decision curve analysis (DCA) showed that patients could potentially benefit more from treatment using the DL-SVM model, and calibration curves demonstrated good agreement with actual outcomes. The DL model based on portal venous-phase abdominal CT accurately predicts the extent of PM in patients before surgery, which can help maximize the benefits of treatment and optimize the patient's treatment plan.

Quantification of left coronary bifurcation angles and plaques by coronary computed tomography angiography for prediction of significant coronary stenosis: A preliminary study with dual-source CT.

PURPOSE: To evaluate the diagnostic performance of left coronary bifurcation angles and plaque characteristics for prediction of coronary stenosis by dual-source CT. METHODS: 106 patients suspected of coronary artery disease undergoing both coronary computed tomography angiography (CCTA) and invasive coronary angiography (CAG) within three months were included. Left coronary bifurcation angles including the angles between the left anterior descending artery and left circumflex artery (LAD-LCx), left main coronary artery and left anterior descending artery (LM-LAD), left main coronary artery and left circumflex artery (LM-LCx) were measured on CT images. CCTA plaque parameters were calculated by plaque analysis software. Coronary stenosis ≥ 50% by CAG was defined as significant. RESULTS: 106 patients with 318 left coronary bifurcation angles and 126 vessels were analyzed. The bifurcation angle of LAD-LCx was significantly larger in left coronary stenosis ≥ 50% than stenosis < 50%, and significantly wider in the non-calcified plaque group than calcified. Multivariable analyses showed the bifurcation angle of LAD-LCx was an independent predictor for significant left coronary stenosis (OR = 1.423, P = 0.002). In ROC curve analysis, LAD-LCx predicted significant left coronary stenosis with a sensitivity of 66.7%, specificity of 78.4%, positive predictive value of 85.2% and negative predictive value of 55.8%. The lipid plaque volume improved the diagnostic performance of CCTA diameter stenosis (AUC: 0.854 vs. 0.900, P = 0.045) in significant coronary stenosis. CONCLUSIONS: The bifurcation angle of LAD-LCx could predict significant left coronary stenosis. Wider LAD-LCx is related to non-calcified lesions. Lipid plaque volume could improve the diagnostic performance of CCTA for coronary stenosis prediction.