高级迭代重建算法降低腹部CT剂量的潜能:体模研究

Dose reduction potential in abdominal CT with advanced iterative reconstruction algorithms: A phantom study

目的探讨将高级迭代重建算法[基于模型的迭代重建(MBIR)技术和自适应统计迭代重建(ASiR)]技术用于降低腹部CT扫描剂量的可行性。方法应用宝石能谱CT(Discovery CT750HD)以不同管电流(400、350、300、250、200、180、160、140、120、100、80、60、50、40、30、20、10mA)对Fluke Biomedical RANDO标准男性模体进行扫描,管电压为均120kV,X线球管旋转时间0.60s,螺距0.984,层厚5mm,层间距5mm,矩阵512×512,DFOV 35cm。记录不同管电流扫描条件下的CT容积剂量指数(CTDIvol)和剂量长度乘积(DLP)。分别用滤过反投影重建(FBP)、50%自适应迭代重建算法(50%ASiR)及模型基础的迭代重建技术(MBIR)进行图像重建,重建层厚均为0.625mm。测量三种重建模式下图像的平均CT值、噪声及对比噪声比(CNR,腰椎与软组织的对比)。结果相同管电流条件下三种重建模式的噪声、CNR差异均有统计学意义(P均<0.05)。不同管电流(400～10mA)条件下,50%ASiR及MBIR重建算法(相对于FBP算法)使噪声分别减少(27.86%～31.46%)及(45.36%～86.37%),SNR分别提高(28.68%～31.08%)及(46.43%～84.38%)。图像能够符合诊断要求的最小管电流分别为FBP:200mA、50%ASiR:140mA及MBIR:80mA。在图像质量类似的情况下,MBIR及50%ASiR模式分别可减少59.91%及35.94%剂量。三种重建模式CT值差异均无统计学意义(P均>0.05)。结论高级重建算法能够减少图像噪声及提高图像CNR,同时具有减少腹部CT扫描剂量的潜能;相对于FBP,MBIR重建算法能够减少约60%的扫描剂量。

Objective To assess the dose reduction potential in abdominal CT of advanced reconstruction algorithms （model-based iterative reconstruction [MBIR] and adaptive statistical iterative reconstruction [ASiR] with a standard man phantom by comparing image noise and contrast-noise-ratio （CNR） with the filtered back projection （FBP） reconstructions. Methods A Fluke Biomedical RANDO standard man phantom was scanned with Discovery CT750 HD under different tube currents （400 mA, 350 mA, 300 mA, 250 mA, 200 mA, 180 mA, 160 mA, 140 mA, 120 mA, 100 mA, 80 mA, 60 mA, 50 mA, 40 mA, 30 mA, 20 mA and 10 mA）, and the other parameters were： Tube voltage was 120 kV, scanning time was 0.60 s, screw pitch was 0. 984, slice thickness was 5 mm, interlamellar spacing was 5 mm, matrix was 512×512, and DFOV was 35 cm. Volume CT dose indexes （CTDIvol） and dose-length product （DLP） were recorded. Three sets of 0. 625 mm slice thickness CT images were reconstructed with FBP, 50~ASiR and MBIR. CT value, imaging noise and contrast-to-noise ratio （CNR） for the lumbar vertebra relative to muscle were measured. Three experienced radiologists as- sessed all the 3 set of images blindly and subjectively. Results Under different tube currents （400--10 mA）, compared with FBP, the average image noise reduction with 50% ASiR and MBIR was （27. 86%--31.46% ） and （45.36%-- 86.37%）, respectively, and the CNR increase for lumbar vertebra with 50MASiR and MBIR was （28.68%--31.08%） and （46.43%--84. 38%）, respectively. The minimum tube current for displaying lumbar vertebra clearly was 200 mA （FBP）, 140 mA （50%ASiR） and 80 mA （MBIR）, respectively. On the other hand, to get similar quality images, MBIR and 50%ASiR could reduce dose by 59.91% and 35.94% compared with FBP, respectively. Conclusion Advanced recon- struction algorithms can greatly reduce image noise and improve image CNR. Compared with FBP, it is possible to reduce dose by 60M in abdominal CT examination with MBIR reconstruction.