双能量CT非线性融合和噪声优化的虚拟单能量图像技术在喉鳞状细胞癌中的应用

Application of dual-energy CT nonlinear blending and noise-optimized virtual monoenergetic imaging in laryngeal squamous cell carcinoma

摘要:

目的 探讨双能量CT线性融合图像（linear blending imaging，LBI）、非线性融合图像（nonlinear blending image，NBI）和噪声优化的虚拟单能量图像（noise-optimized virtual monoenergetic image，VMI+）技术在喉鳞状细胞癌中的应用价值。方法 回顾性分析2019年6月至2022年3月61例经病理证实为喉鳞状细胞癌患者的双能量CT资料。双能量图像采用LBI［融合系数为1（80 kV）和0.6（M0.6）］、NBI和VMI+（40 keV、55 keV）技术重建。比较5组图像的客观图像质量［对比噪声比（contrast-to-noise ratio，CNR）、肿瘤CT值、噪声］和主观图像质量（肿瘤边界评分和整体图像质量评分）。结果 40 keV的CNR、肿瘤CT值和肿瘤边界评分均明显高于80 kV、M0.6、NBI和55 keV，差异均有统计学意义（均P<0.05）。NBI的整体图像质量评分明显高于80 kV、M0.6、40 keV和55 keV，差异均有统计学意义（均P<0.05）。NBI的噪声明显低于80 kV、40 keV和55 keV，差异均有统计学意义（均P<0.05）。结论 在喉鳞状细胞癌的双能量CT中，采用VMI+技术（40 keV）能提供更好的CNR、肿瘤CT值和肿瘤边界，采用NBI技术能提供更低的噪声和更好的整体图像质量。

Abstract:

Objective To investigate the application value of dual-energy CT（DECT） linear blending imaging（LBI），nonlinear blending imaging（NBI），and noise-optimized virtual monoenergetic imaging（VMI+） techniques in laryngeal squamous cell carcinoma （SCC）.Methods A retrospective analysis was performed for the DECT data of 61 patients with pathologically confirmed laryngeal SCC from June 2019 to March 2022. DECT images were reconstructed using LBI with weighting factors 1 （80 kV） and 0.6（M0.6），NBI，and VMI+（40 keV and 55 keV） techniques. The five groups of images were compared in terms of objective image quality ［contrast-to-noise ratio （CNR），tumor CT value，and noise］ and subjective image quality（tumor delineation score and overall image quality score）.Results The 40 keV images had significantly higher CNR，tumor CT value，and tumor delineation score than 80 kV，M0.6，NBI，and 55 keV images（all P<0.05）. The NBI images had a significantly higher overall image quality score than the 80 kV，M0.6，40 keV，and 55 keV images（all P<0.05）. The NBI images had a significantly lower noise than the 80 kV，40 keV，and 55 keV images（all P<0.05）.Conclusion In DECT for laryngeal SCC，the VMI+（40 keV） technique can provide images with better CNR，tumor CT value，and tumor delineation，while the NBI technique can provide images with lower noise and better overall image quality.