Using dual-layer detector spectral computed tomography to predict lymph node metastasis and identify differentiation degree of gastric cancer

Abstract

Purpose: To investigate the value of spectral computed tomography (CT) in evaluating the differentiation degree of gastric cancer and predicting lymph node (LN) metastasis. Methods: Enhanced spectral CT images of 51 gastric cancer patients were retrospectively analyzed. Effective atomic number (Zeff), normalized effective atomic number (nZeff), iodine density (ID), and normalized ID (nID) of the gastric cancer were measured at the venous stage. Zeff, nZeff, ID and nID of the study groups (poorly and moderately-highly differentiated groups; LN metastasis and no LN metastasis groups) were statistically compared using independent t-tests. Sex, age, location, and thickness of gastric cancer were analyzed using one-way analysis of variance or independent t-test. Correlations were assessed using Spearman rank correlation. The optimal diagnostic cutoff value with sensitivity, specificity, and area under the curve (AUC) was determined using receiver operating characteristic curve analysis.Results: Zeff, nZeff, ID and nID of gastric cancer were higher in the poorly differentiated group than in the moderately-highly differentiated group (P<0.05), and AUCs for evaluating the degree of differentiation were 0.909, 0.888, 0.905, and 0.916, respectively. There were no significant differences in Zeff, nZeff, ID and nID regarding age, sex, tumor thickness, or tumor location (P>0.05). Zeff, nZeff, ID and nID were higher in the LNM group than in the no LNM group (P<0.05), and AUCs for predicting LNM were 0.888, 0.824, 0.828 and 0.892, respectively. Conclusions: Zeff, nZeff, ID and nID values of spectral CT predicted the degree of differentiation of gastric cancer and LNM and were positively correlated with N stage.

A nomogram model of spectral CT quantitative parameters and clinical characteristics predicting lymphovascular invasion of gastric cancer

“…Therefore, more comprehensive data and multiple indicators were required to construct the predictive model with the optimal diagnostic efficiency. It is widely known that spectral CT quantitative parameters can predict lymph node metastasis in GC [ 26 , 27 ], which was consistent with our finding that some spectral quantitative parameters were highly correlated with LVI. As a result, VIC was an independent risk factor of LVI increased the probability of LVI by 1.165-fold for every unit increase in its value, which was consistent with the previous study [ 2 ].…”