Prediction of histopathological grade in gastric cancer using dual energy CT-derived synthetic extracellular volume fraction

Purpose To investigate the relationship between the synthetic extracellular volume (ECV) fraction and the hematocrit (HCT)-based ECV fraction in gastric cancer patients and to evaluate its potential utility in predicting the pathological grade of gastric cancer. Methods Data from the derivation cohort consisting of 142 patients who underwent nonenhanced abdominal CT examination were collected, and the CT values of the abdominal aorta (CTA) and inferior vena cava (CTV) were measured separately. Using linear regression between CT attenuations and conventional hematocrit (HCT con ), calculation formulas were derived for the synthetic hematocrit of the abdominal aorta (HCT syn, A ) and the inferior vena cava (HCT syn, V ). A validation cohort of 62 gastric adenocarcinoma patients with dual-energy CT was analyzed to calculate synthetic ECV (ECV syn, A and ECV syn, V ) using the derived formulas. Differences between synthetic and conventional ECV(ECVcon), and their efficacy in differentiating pathological grades, were compared. Results Derived formulas: HCT syn,A% = 1.152× CT A - 12.311; HCT syn,V% = 1.142× CT V - 11.229. In validation, HCT syn,A and HCT syn,V showed no significant difference from HCT con (all P>0.05). HCT con was strongly correlated with HCT syn, A (r = 0.898, P < 0.01) and HCT syn, V (r = 0.826, P < 0.01). ECV syn,A and ECV syn,V also did not differ from ECV con (all P>0.05). ECV con was strongly correlated with ECV syn, A (r = 0.990, P < 0.01) and ECV syn, V (r = 0.983, P < 0.01). All ECV metrics exhibited excellent and comparable efficiency in differentiating pathological grades. Conclusions Synthetic ECV performs comparably to conventional ECV and predicts gastric cancer pathological grading effectively.