ISSN 1004-4140
CN 11-3017/P

深度学习重建算法的图像质量体模研究

刘方韬, 刘隺是, 陈勇, 姜江, 王凌云, 董海鹏, 张勇, 张璇, 孔德艳, 常蕊

刘方韬, 刘隺是, 陈勇, 等. 深度学习重建算法的图像质量体模研究[J]. CT理论与应用研究, 2022, 31(3): 351-356. DOI: 10.15953/j.ctta.2021.061.
引用本文: 刘方韬, 刘隺是, 陈勇, 等. 深度学习重建算法的图像质量体模研究[J]. CT理论与应用研究, 2022, 31(3): 351-356. DOI: 10.15953/j.ctta.2021.061.
LIU F T, LIU H S, CHEN Y, et al. Image quality assessment for deep learning image reconstruction algorithm: A phantom study[J]. CT Theory and Applications, 2022, 31(3): 351-356. DOI: 10.15953/j.ctta.2021.061. (in Chinese).
Citation: LIU F T, LIU H S, CHEN Y, et al. Image quality assessment for deep learning image reconstruction algorithm: A phantom study[J]. CT Theory and Applications, 2022, 31(3): 351-356. DOI: 10.15953/j.ctta.2021.061. (in Chinese).

深度学习重建算法的图像质量体模研究

详细信息
    作者简介:

    刘方韬: 男,上海交通大学附属瑞金医院放射科主管技师、技师长助理,主要从事放射技术工作,擅长CT、MR影像学科研工作,E-mail:lft40620@rjh.com.cn

    刘隺是: 男,上海交通大学医学院附属瑞金医院放射科主管技师,主要从事医学图像采集及重建工作,E-mail:lhs40432@rjh.com.cn

  • 中图分类号: O  242;TP  391.41

Image Quality Assessment for Deep Learning Image Reconstruction Algorithm: A Phantom Study

  • 摘要: 目的:使用体模比较CT深度学习重建算法和迭代重建算法的图像质量。方法:使用GE Revolution Apex扫描ACR质量控制体模Gammex 464,分别测量module 1~module 4的5种物质CT值准确性、低对比度分辨率、图像均匀性和高对比度分辨率。通过指标比较高剂量下(20 mGy)深度学习重建算法TrueFedelity<sup<TM</sup<(TFI)3种等级(DL、DM及DH)和自适应统计迭代重建算法V(AV)3种等级(30%、60% 及90%)的图像质量。两种算法的各指标比较采用单因素方差分析。结果:所有6组图像的高/低分辨率均一致(高对比度分辨率:10 lp/cm;低对比度分辨率:6 mm);两种算法都轻微高估聚乙烯、空气以及丙烯酸的CT值,各物质间CT值差异不具有统计学意义。两种算法均低估骨和固态水的CT值,其中,TFI算法对固态水的CT值较AV更接近真实值,但各组图像间不具有统计学差异。6组图像中,TFIDH的图像均匀性最佳;同等级条件下,深度学习重建算法相较IR算法的图像均匀性更佳。结论:深度学习重建算法在高剂量水平下可以在保持图像空间分辨率和CT值准确性的基础上,进一步降低图像噪声。
    Abstract: Objective: To compare the image quality between the deep learning image reconstruction (DLIR) and iterative reconstruction (IR) algorithms via the dedicated phantom. Method: ACR quality assurance phantom (Gammex 464) was scanned by GE Revolution Apex. The CT value accuracy, low contrast resolution, image uniformity and high contrast resolution of five substances from module 1 to module 4 were measured respectively. Through the above indicators, the image quality of three levels (DL, DM and DH) of Truefedelitytm (TFI) and three levels (30%, 60% and 90%) of adaptive statistical iterative reconstruction-V (asir-V, hereinafter referred to as AV) under high dose (20 mgy) were compared. The comparison between the two algorithms for each parameter was tested by One-Way Anova. Results: The high/low-contrast resolution of the six image series were consistent (high-contrast resolution: 10 lp/cm; low-contrast resolution: 6 mm). The two algorithms both slightly overestimated the CT value of polyethylene, air and acrylic, and no statistically significant difference was found among the difference of CT values of the substances. Both algorithms underestimated the CT value of bone and the solid water; TFI showed better performance in evaluating the solid water which was closer to the real value, though statistical difference was not found between each group of images. Among the 6 groups of images, TFI DH showed the best image uniformity and at the same reconstruction level, TFI showed better uniformity than AV. Conclusion: DLIR can further reduced image noise while maintaining image spatial resolution and CT value accuracy at high dose level.
  • 图  1   ACR Gammex 464体模

    Figure  1.   The ACR Gammex 464 phantom

    表  1   各组图像5种物质CT值与图像均匀性测量结果

    Table  1   The measurment results of CT values of the five materials and image uniformnity

    物质/指标A组(AV 30%)B组(AV 60%)C组(AV 90%)D组(TFI DL)E组(TFI DM)F组(TFI DH)P
    聚乙烯-94.9±0.3-95.0±0.3-94.9±0.3-95.0±0.4-94.9±0.3-94.9±0.30.994
    881.8±0.7881.8±0.6881.7±0.6881.9±0.7881.9±0.7881.8±0.80.999
    空气-994.1±0.8-994.3±0.7-994.3±0.6-994.3±0.7-994.3±0.6-994.4±0.50.996
    丙烯酸121.5±0.8121.6±0.4121.6±0.4121.1±0.7121.1±0.5121.1±0.40.640
    固态水-2.1±1.0-2.0±1.0-2.2±1.2-1.9±0.7-1.7±0.6-1.7±0.70.979
    均匀性 4.0 3.7 2.9 3.8 3.3 2.7
     注:聚乙烯、骨、空气、丙烯酸和固态水的理论值分别为 -95、950、-1000 、120和0 HU,图像均匀性的单位是HU。
    下载: 导出CSV
  • [1]

    BEREGI J P, GREFFIER J. Low and ultra-low dose radiation in CT: Opportunities and limitations[J]. Diagnostic and Interventional Imaging, 2019, 100(2): 63−64. DOI: 10.1016/j.diii.2019.01.007.

    [2]

    MACRI F, GREFFIER J, KHASANOVA E, et al. Minor blunt thoracic trauma in the emergency department: Sensitivity and specificity of chest ultralow-dose computed tomography compared with conventional radiography[J]. Annals of Emergency Medicine, 2019, 73(6): 665−670. DOI: 10.1016/j.annemergmed.2018.11.012.

    [3]

    KIM H G, LEE H J, LEE S K, et al. Head CT: Image quality improvement with ASIR-V using a reduced radiation dose protocol for children[J]. European Radiology, 2017, 27(9): 3609−3617. DOI: 10.1007/s00330-017-4733-z.

    [4]

    LARBI A, ORLIAC C, FRANDON J, et al. Detection and characterization of focal liver lesions with ultra-low dose computed tomography in neoplastic patients[J]. Diagnostic and Interventional Imaging, 2018, 99(5): 311−320. DOI: 10.1016/j.diii.2017.11.003.

    [5]

    TANG H, LIU Z, HU Z, et al. Clinical value of a new generation adaptive statistical iterative reconstruction (ASIR-V) in the diagnosis of pulmonary nodule in low-dose chest CT[J]. British Journal of Radiology, 2019, 92(1103): 20180909. DOI: 10.1259/bjr.20180909.

    [6] 仵腾辉, 查云飞, 杨峰. 不同螺距联合ASIR重建技术在COVID-19胸部低剂量CT扫描中的应用研究[J]. CT理论与应用研究, 2022,31(2): 194−201. DOI: 10.15953/j.1004-4140.2022.31.02.05.

    WU T H, ZHA Y F, YANG F. The application and study of different pitch combined with ASIR in low-dose chest CT screening on COVID-19[J]. CT Theory and Applications, 2022, 31(2): 194−201. DOI: 10.15953/j.1004-4140.2022.31.02.05. (in Chinese).

    [7]

    VERDUN F R, RACINE D, OTT J G, et al. Image quality in CT: From physical measurements to model observers[J]. Physica Medica, 2015, 31(8): 823−843. DOI: 10.1016/j.ejmp.2015.08.007.

    [8]

    OTT J G, BECCE F, MONNIN P, et al. Update on the non-prewhitening model observer in computed tomography for the assessment of the adaptive statistical and model-based iterative reconstruction algorithms[J]. Physics in Medicine and Biology, 2014, 59(15): 4047−4064. DOI: 10.1088/0031-9155/59/4/4047.

    [9]

    SAMEI E, BAKALYAR D, BOEDEKER K L, et al. Performance evaluation of computed tomography systems: Summary of AAPM task group 233[J]. Medical Physics, 2019, 46(11): e735−e756. DOI: 10.1002/mp.13763.

    [10]

    GEYER L L, SCHOEPF U J, MEINEL F G, et al. State of the art: Iterative CT reconstruction techniques[J]. Radiology, 2015, 276(2): 339−357. DOI: 10.1148/radiol.2015132766.

    [11]

    GREFFIER J, HAMARD A, PEREIRA F, et al. Image quality and dose reduction opportunity of deep learning image reconstruction algorithm for CT: A phantom study[J]. European Radiology, 2020, 30(7): 3951−3959. DOI: 10.1007/s00330-020-06724-w.

    [12]

    LYU P, NEELY B, SOLOMON J, et al. Effect of deep learning image reconstruction in the prediction of resectability of pancreatic cancer: Diagnostic performance and reader confidence[J]. European Journal of Radiology, 2021, 141: 109825. DOI: 10.1016/j.ejrad.2021.109825.

    [13]

    FRANCK C, ZHANG G, DEAK P, et al. Preserving image texture while reducing radiation dose with a deep learning image reconstruction algorithm in chest CT: A phantom study[J]. Physica Medica, 2021, 81: 86−93. DOI: 10.1016/j.ejmp.2020.12.005.

    [14]

    GREFFIER J, FRANDON J, Si-MOHAMED S, et al. Comparison of two deep learning image reconstruction algorithms in chest CT images: A task-based image quality assessment on phantom data[J]. Diagnostic and Interventional Imaging, 2021, S2211-5684(21): 00174−1. DOI: 10.1016/j.diii.2021.08.001.

    [15] 温德英, 杨杰尹, 汪琴, 等. 深度学习重建算法在上腹部CT成像中的应用[J]. CT理论与应用研究, 2021,31(3): 329−336. DOI: 10.15953/j.ctta.2021-005.

    WEN D Y, YANG J Y, WANG Q, et al. Application of deep learning reconstruction algorithm in upper abdomen CT[J]. CT Theory and Applications, 2021, 31(3): 329−336. DOI: 10.15953/j.ctta.2021-005. (in Chinese).

    [16]

    ICHIKAWA Y, KANII Y, YAMAZAKI A, et al. Deep learning image reconstruction for improvement of image quality of abdominal computed tomography: Comparison with hybrid iterative reconstruction[J]. Japanese Journal of Radiology, 2021, 39(6): 598−604. DOI: 10.1007/s11604-021-01089-6.

    [17]

    PARK C, CHOO K S, JUNG Y, et al. CT iterative vs deep learning reconstruction: Comparison of noise and sharpness[J]. European Radiology, 2021, 31(5): 3156−3164. DOI: 10.1007/s00330-020-07358-8.

图(1)  /  表(1)
计量
  • 文章访问数:  377
  • HTML全文浏览量:  68
  • PDF下载量:  26
  • 被引次数: 0
出版历程
  • 收稿日期:  2021-12-05
  • 录用日期:  2022-01-11
  • 网络出版日期:  2022-01-20
  • 发布日期:  2022-05-22

目录

    /

    返回文章
    返回
    x 关闭 永久关闭