论文部分内容阅读
目的比较不同水平的自适应统计迭代重建(ASiR)算法在门静脉成像中的图像质量,优化门静脉成像中的ASiR重建算法的比例。方法搜集临床有恶性肿瘤病史、治疗后随诊,行上腹部增强扫描的患者40例。采用能谱模式扫描,双电压快速切换,从扫描数据中重组出60 ke V能量段的单能量图像,分别用FBP(0%ASiR)、30%ASiR、40%ASiR、50%ASiR、60%ASiR算法进行重建,得到5组不同水平ASiR重建算法的图像。采用单因素方差分析比较5组图像中门静脉主干、肝实质的客观噪声。由两名腹部影像诊断医师在盲法下对5组图像的噪声、伪影及门静脉主干第3级分支细小结构的显示进行主观图像质量评分。结果与FBP(0%ASiR)比较,30%ASiR、40%ASiR、50%ASiR、60%ASiR图像中,门静脉主干及肝实质的客观噪声值依次减低,且差异均有统计学意义(P均=0.0001)。40%ASiR组分别与50%ASiR、60%ASiR组之间比较,门静脉主干及肝实质的客观噪声值差异均无统计学意义(P均>0.0501)。随着ASiR值的升高,两名医师在盲法下对5种不同重建算法图像的噪声、伪影主观评分均依次升高,仅显示门静脉主干第3级分支。主观评分中40%ASiR图像最高,60%ASiR图像最低。5种不同重建算法之间图像噪声、伪影、显示门静脉主干第3级分支结构主观评分的差异均有统计学意义(P均=0.000)。在门静脉主干第3级分支结构的显示方面,40%ASiR图像主观评分分别高于50%ASiR、60%ASiR图像,且差异均有统计学意义(P=0.000、0.000)。结论采用ASiR算法可通过降低图像噪声来提高图像质量,随着ASiR重建比例的升高,图像噪声、伪影依次减少,图像密度分辨率依次增高。但过高的ASiR算法本身会引起图像的“模糊”效应。在门静脉成像中,将ASiR设置为40%时,既可以降低图像噪声,又不会对图像产生明显的“模糊”效应,图像质量最佳。
Objective To compare the image quality of different levels of adaptive statistical iterative reconstruction (ASiR) algorithm in portal vein imaging and to optimize the proportion of ASiR reconstruction in portal vein imaging. Methods A total of 40 patients with history of malignant tumor, follow-up after treatment and contrast-enhanced abdominal scan were collected. Single energy images of 60 ke V energy segments were reconstructed from the scan data using spectral mode scanning and dual voltage fast switching. The single energy images were reconstructed with FBP (0% ASiR), 30% ASiR, 40% ASiR, 50% ASiR, ASiR algorithm to reconstruct the image of 5 different ASiR reconstruction algorithms. One-way ANOVA was used to compare the objective noise of portal vein and liver parenchyma in five groups of images. Subjective image quality scores were scored by two abdominal imaging diagnosticians under the blinded method for the noise and artifacts of the five groups of images and the display of the fine structure of the third branch of the portal vein trunk. Results Compared with FBP (0% ASiR), the objective noise values of portal vein and liver parenchyma decreased in 30% ASiR, 40% ASiR, 50% ASiR and 60% ASiR images, respectively, with statistical significance = 0.0001). There were no significant differences in the objective noise values between the 40% ASiR group and the 50% ASiR and 60% ASiR groups (P> 0.0501). With the increase of ASiR, the subjective scores of the noise and artifacts of the five doctors under the blind method increased successively, showing only the level 3 branch of the main portal vein. Subjective scores in 40% ASiR image highest, 60% ASiR image lowest. The image noise and artifacts among the five different reconstruction algorithms showed that there were significant differences in the subjective scores of the grade 3 branches of portal vein (P = 0.000). The subjective scores of 40% ASiR images were higher than that of 50% ASiR and 60% ASiR images in the display of the third branch of portal vein. The difference was statistically significant (P = 0.000, 0.000). Conclusion The ASiR algorithm can improve the image quality by reducing the image noise. As the ASiR reconstruction ratio increases, the image noise and artifacts decrease in turn, and the image density resolution increases successively. But the overly high ASiR algorithm can cause the “fuzzy” effect of the image. In portal vein imaging, setting the ASiR to 40% reduces image noise without significant “blurring” effects on the image, resulting in the best image quality.