| 摘要: |
| 目的：探讨双层探测器光谱CT多参数预测结直肠癌（CRC）神经侵犯（PNI）中的临床价值。
方法：回顾性收集103例最终经病理证实的结直肠癌患者，根据 PNI结果将其分为PNI（+）组( n =59 ) 和PNI（-）组( n = 44) 。所有患者均行皓克Spectral CT全腹部平扫+双期增强（动脉期、静脉期），测量动脉期、静脉期病灶CT40 keV、CT70 keV、CT90 keV、碘浓度（IC）、有效原子序数（Zeff）、电子云密度（ED）及同层面大动脉的CT值，并计算标准化碘浓度（NIC），能谱曲线斜率（λH），比较PNI（+）组与PNI（-）组各组间能谱参数的差异。将差异有统计学意义的纳入二元logistic回归分析中，再构建能谱多参数预测模型，并绘制受试者工作（ROC）曲线评估各项参数及联合模型对PNI的诊断效能。
结果：PNI（+）组和PNI（-）组病灶之间，动脉期病灶IC、NIC、Zeff、λHU及CT40keV、CT70keV，静脉期病灶IC、NIC、λHU、CT40keV、CT90keV差异有统计学意义（P均＜0.05），其余光谱参数差异均无统计学意义（P均＞0.05）。logistic 回归结果显示动脉期病灶λHU、静脉期病灶IC、病灶NIC为预测结直肠癌 PNI的独立因素（P 均＜0.05，），其单独预测结直肠癌PNI的 AUC 分别为0.731、0.614及 0.550 ，以此参数构建联合模型其诊断敏感度为80.4%，特异度为79.5%，AUC 为 0.845。
结论：基于双层探测器光谱CT多参数可在术前预测CRC-PNI，为临床提供一种新的术前评估方式。 |
| 关键词: 光谱CT 结直肠癌 神经侵犯 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| Spectral CT multi-parameter quantitative prediction of perineural invasion in colorectal cancer |
|
FU Lanqi, WANG Junna, XU Zhichao, WANG Shiwei
|
|
The First Affiliated Hospital of Zhejiang Chinese Medical University (Zhejiang Provincial Hospital of Traditional Chinese Medicine)
|
| Abstract: |
| Objective:?To evaluate the clinical value of dual-layer detector spectral CT multiparameters in predicting perineural invasion (PNI) in colorectal cancer (CRC).
Methods: We retrospectively analyzed 103 pathologically confirmed CRC patients, stratified into PNI-positive (n=59) and PNI-negative (n=44) groups.
All patients underwent non-contrast and dual-phase contrast-enhanced (arterial phase, venous phase) abdominal scans using the HAWK Spectral CT system. The following spectral parameters of the lesions were measured in both arterial and venous phases: CT values at 40 keV, 70 keV, and 90 keV; iodine concentration (IC); effective atomic number (Zeff); electron density (ED); as well as the CT values of the aorta at the same level. The normalized iodine concentration (NIC) and spectral curve slope (λHU) were subsequently calculated. Differences in spectral parameters between the PNI (+) group and PNI (-) group were then compared. Diagnostic performance was assessed using receiver operating characteristic (ROC) curve analysis.
Results:Significant differences (all P<0.05) were observed in: arterial-phase IC, NIC, Zeff, λHU, CT40keV, and CT70keV; venous-phase IC, NIC, λHU, CT40keV, and CT90keV. Logistic regression identified arterial-phase λHU, venous-phase IC, and NIC as independent PNI predictors (all P<0.05), with individual AUCs of 0.731, 0.614, and 0.550 respectively. The combined model achieved 80.4% sensitivity, 79.5% specificity, and 0.845 AUC.
Conclusion: Dual-layer spectral CT multiparameter analysis enables reliable preoperative prediction of CRC PNI, offering a novel clinical assessment tool. |
| Key words: Spectral CT Colorectal cancer Perineural invasion |