Scholarlink

Purpose <p>This study aimed to develop and validate a nomogram integrating dual-energy computed tomography (DECT)-derived extracellular volume fraction (ECV), spectral quantitative parameters, morphological features, and clinical variables for predicting early recurrence (ER) after gastrectomy in patients with gastric cancer (GC).</p> Materials and methods <p>This retrospective study included GC patients from three institutions. Two radiologists independently evaluated ECV, spectral quantitative parameters, and morphological features. Independent predictors were identified through logistic regression and used to construct a nomogram, DECT, conventional CT, and clinical models. The predictive performance of the nomogram was assessed using calibration curves, receiver operating characteristic (ROC) analysis, and decision curve analysis (DCA). The association between nomogram-predicted ER risk and recurrence-free survival (RFS) was evaluated using Kaplan–Meier survival analysis and Cox regression. A total of 354 patients were included in the study and stratified into a training cohort (<i>n</i> = 195), an internal validation cohort (<i>n</i> = 84), and an external validation cohort (<i>n</i> = 75).</p> Results <p>ECV, arterial enhancement fraction (AEF), CT–detected extramural venous invasion (CT-EMVI), and neutrophil-to-lymphocyte ratio (NLR) were independently associated with ER. The nomogram yielded areas under the ROC curve (AUCs) of 0.894, 0.865, and 0.875 in three cohorts, outperforming the DECT, conventional CT, and clinical models. The nomogram showed good agreement between predicted and observed outcomes in all cohorts, as evidenced by non-significant Hosmer–Lemeshow test results (<i>P</i> &gt; 0.05), and yielded favorable clinical utility according to decision curve analysis. Nomogram-predicted ER risk was independently correlated with RFS, and risk stratification based on the nomogram successfully identified high-risk patients with significantly worse 4-year RFS probabilities in all cohorts.</p> Conclusion <p>A nomogram incorporating ECV, AEF, CT-EMVI, and NLR provided an effective tool for predicting ER after gastrectomy in patients with GC.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dual-energy computed tomography-derived extracellular volume fraction and spectral quantitative parameters for predicting early recurrence after gastrectomy: insights from a multicenter study

  • 期刊论文

  • 发表日期：

    2026年6月26日

查看全文
  • Zhengqi Zhu,
  • Anyi Song,
  • Jian Shi,
  • Aiyan Fu,
  • Ru Zhang,
  • Mengyue Zhang,
  • Mimi Mao

摘要

Purpose

This study aimed to develop and validate a nomogram integrating dual-energy computed tomography (DECT)-derived extracellular volume fraction (ECV), spectral quantitative parameters, morphological features, and clinical variables for predicting early recurrence (ER) after gastrectomy in patients with gastric cancer (GC).

Materials and methods

This retrospective study included GC patients from three institutions. Two radiologists independently evaluated ECV, spectral quantitative parameters, and morphological features. Independent predictors were identified through logistic regression and used to construct a nomogram, DECT, conventional CT, and clinical models. The predictive performance of the nomogram was assessed using calibration curves, receiver operating characteristic (ROC) analysis, and decision curve analysis (DCA). The association between nomogram-predicted ER risk and recurrence-free survival (RFS) was evaluated using Kaplan–Meier survival analysis and Cox regression. A total of 354 patients were included in the study and stratified into a training cohort (n = 195), an internal validation cohort (n = 84), and an external validation cohort (n = 75).

Results

ECV, arterial enhancement fraction (AEF), CT–detected extramural venous invasion (CT-EMVI), and neutrophil-to-lymphocyte ratio (NLR) were independently associated with ER. The nomogram yielded areas under the ROC curve (AUCs) of 0.894, 0.865, and 0.875 in three cohorts, outperforming the DECT, conventional CT, and clinical models. The nomogram showed good agreement between predicted and observed outcomes in all cohorts, as evidenced by non-significant Hosmer–Lemeshow test results (P > 0.05), and yielded favorable clinical utility according to decision curve analysis. Nomogram-predicted ER risk was independently correlated with RFS, and risk stratification based on the nomogram successfully identified high-risk patients with significantly worse 4-year RFS probabilities in all cohorts.

Conclusion

A nomogram incorporating ECV, AEF, CT-EMVI, and NLR provided an effective tool for predicting ER after gastrectomy in patients with GC.