DOI QR코드

DOI QR Code

Validation of Deep-Learning Image Reconstruction for Low-Dose Chest Computed Tomography Scan: Emphasis on Image Quality and Noise

  • Joo Hee Kim (Department of Radiology, Veterans Health Service Medical Center) ;
  • Hyun Jung Yoon (Department of Radiology, Veterans Health Service Medical Center) ;
  • Eunju Lee (Department of Radiology, Veterans Health Service Medical Center) ;
  • Injoong Kim (Department of Radiology, Veterans Health Service Medical Center) ;
  • Yoon Ki Cha (Department of Radiology, Dongguk University Ilsan Hospital) ;
  • So Hyeon Bak (Department of Radiology, Kangwon National University Hospital, Kangwon National University School of Medicine)
  • 투고 : 2020.02.12
  • 심사 : 2020.05.18
  • 발행 : 2021.01.01

초록

Objective: Iterative reconstruction degrades image quality. Thus, further advances in image reconstruction are necessary to overcome some limitations of this technique in low-dose computed tomography (LDCT) scan of the chest. Deep-learning image reconstruction (DLIR) is a new method used to reduce dose while maintaining image quality. The purposes of this study was to evaluate image quality and noise of LDCT scan images reconstructed with DLIR and compare with those of images reconstructed with the adaptive statistical iterative reconstruction-Veo at a level of 30% (ASiR-V 30%). Materials and Methods: This retrospective study included 58 patients who underwent LDCT scan for lung cancer screening. Datasets were reconstructed with ASiR-V 30% and DLIR at medium and high levels (DLIR-M and DLIR-H, respectively). The objective image signal and noise, which represented mean attenuation value and standard deviation in Hounsfield units for the lungs, mediastinum, liver, and background air, and subjective image contrast, image noise, and conspicuity of structures were evaluated. The differences between CT scan images subjected to ASiR-V 30%, DLIR-M, and DLIR-H were evaluated. Results: Based on the objective analysis, the image signals did not significantly differ among ASiR-V 30%, DLIR-M, and DLIR-H (p = 0.949, 0.737, 0.366, and 0.358 in the lungs, mediastinum, liver, and background air, respectively). However, the noise was significantly lower in DLIR-M and DLIR-H than in ASiR-V 30% (all p < 0.001). DLIR had higher signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) than ASiR-V 30% (p = 0.027, < 0.001, and < 0.001 in the SNR of the lungs, mediastinum, and liver, respectively; all p < 0.001 in the CNR). According to the subjective analysis, DLIR had higher image contrast and lower image noise than ASiR-V 30% (all p < 0.001). DLIR was superior to ASiR-V 30% in identifying the pulmonary arteries and veins, trachea and bronchi, lymph nodes, and pleura and pericardium (all p < 0.001). Conclusion: DLIR significantly reduced the image noise in chest LDCT scan images compared with ASiR-V 30% while maintaining superior image quality.

키워드

참고문헌

  1. Dela Cruz CS, Tanoue LT, Matthay RA. Lung cancer: epidemiology, etiology, and prevention. Clin Chest Med 2011;32:605-644 https://doi.org/10.1016/j.ccm.2011.09.001
  2. National Lung Screening Trial Research Team; Aberle DR, Adams AM, Berg CD, Black WC, Clapp JD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011;365:395-409 https://doi.org/10.1056/NEJMoa1102873
  3. National Lung Screening Trial Research Team; Church TR, Black WC, Aberle DR, Berg CD, Clingan KL, et al. Results of initial low-dose computed tomographic screening for lung cancer. N Engl J Med 2013;368:1980-1991 https://doi.org/10.1056/NEJMoa1209120
  4. Diederich S, Wormanns D, Semik M, Thomas M, Lenzen H, Roos N, et al. Screening for early lung cancer with low-dose spiral CT: prevalence in 817 asymptomatic smokers. Radiology 2002;222:773-781 https://doi.org/10.1148/radiol.2223010490
  5. Ohno Y, Koyama H, Seki S, Kishida Y, Yoshikawa T. Radiation dose reduction techniques for chest CT: principles and clinical results. Eur J Radiol 2019;111:93-103 https://doi.org/10.1016/j.ejrad.2018.12.017
  6. Geyer LL, Schoepf UJ, Meinel FG, Nance JW Jr, Bastarrika G, Leipsic JA, et al. State of the art: iterative CT reconstruction techniques. Radiology 2015;276:339-357 https://doi.org/10.1148/radiol.2015132766
  7. Padole A, Ali Khawaja RD, Kalra MK, Singh S. CT radiation dose and iterative reconstruction techniques. AJR Am J Roentgenol 2015;204:W384-W392 https://doi.org/10.2214/AJR.14.13241
  8. Akagi M, Nakamura Y, Higaki T, Narita K, Honda Y, Zhou J, et al. Deep learning reconstruction improves image quality of abdominal ultra-high-resolution CT. Eur Radiol 2019;29:6163-6171 https://doi.org/10.1007/s00330-019-06170-3
  9. Benz DC, Benetos G, Rampidis G, von Felten E, Bakula A, Sustar A, et al. Validation of deep-learning image reconstruction for coronary computed tomography angiography: impact on noise, image quality and diagnostic accuracy. J Cardiovasc Comput Tomogr 2020 Jan 13 [Epub]. https://doi.org/10.1016/j.jcct.2020.01.002
  10. Higaki T, Nakamura Y, Tatsugami F, Nakaura T, Awai K. Improvement of image quality at CT and MRI using deep learning. Jpn J Radiol 2019;37:73-80 https://doi.org/10.1007/s11604-018-0796-2
  11. Liu J, Zhang Y, Zhao Q, Lv T, Wu W, Cai N, et al. Deep iterative reconstruction estimation (DIRE): approximate iterative reconstruction estimation for low dose CT imaging. Phys Med Biol 2019;64:135007
  12. Singh R, Digumarthy SR, Muse VV, Kambadakone AR, Blake MA, Tabari A, et al. Image quality and lesion detection on deep learning reconstruction and iterative reconstruction of submillisievert chest and abdominal CT. AJR Am J Roentgenol 2020;214:566-573 https://doi.org/10.2214/AJR.19.21809
  13. Tatsugami F, Higaki T, Nakamura Y, Yu Z, Zhou J, Lu Y, et al. Deep learning-based image restoration algorithm for coronary CT angiography. Eur Radiol 2019;29:5322-5329 https://doi.org/10.1007/s00330-019-06183-y
  14. Liu P, Wang M, Wang Y, Yu M, Wang Y, Liu Z, et al. Impact of deep learning-based optimization algorithm on image quality of low-dose coronary CT angiography with noise reduction: a prospective study. Acad Radiol 2019 Dec 18 [Epub]. https://doi.org/10.1016/j.acra.2019.11.010
  15. Jensen CT, Liu X, Tamm EP, Chandler AG, Sun J, Morani AC, et al. Image quality assessment of abdominal CT by use of new deep learning image reconstruction: initial experience. AJR Am J Roentgenol 2020;215:50-57 https://doi.org/10.2214/AJR.19.22332
  16. Shin YJ, Chang W, Ye JC, Kang E, Oh DY, Lee YJ, et al. Low-dose abdominal CT using a deep learning-based denoising algorithm: a comparison with ct reconstructed with filtered back projection or iterative reconstruction algorithm. Korean J Radiol 2020;21:356-364 https://doi.org/10.3348/kjr.2019.0413
  17. Hsieh J, Liu E, Nett B, Tang J, Thibault JB, Sahney S. A new era of image reconstruction: TrueFidelityTM: technical white paper on deep learning image reconstruction. Available at: https://pdfs.semanticscholar.org/d0f8/e1e8868e9f8ed22ad5972420139551552e91.pdf?_ga=2.233526110.1531411842.1594709320-2066918258.1594709320. Accessed January 13, 2020
  18. Trattner S, Halliburton S, Thompson CM, Xu Y, Chelliah A, Jambawalikar SR, et al. Cardiac-specific conversion factors to estimate radiation effective dose from dose-length product in computed tomography. JACC Cardiovasc Imaging 2018;11:64-74 https://doi.org/10.1016/j.jcmg.2017.06.006
  19. Christner JA, Braun NN, Jacobsen MC, Carter RE, Kofler JM, McCollough CH. Size-specific dose estimates for adult patients at CT of the torso. Radiology 2012;265:841-847 https://doi.org/10.1148/radiol.12112365
  20. American Association of Physicists in Medicine. Size-specific dose estimates (SSDE) in pediatric and adult body CT examinations. College Park, MD: American Association of Physicists in Medicine, 2011:204
  21. Lin S, Lin M, Lau KK. Image quality comparison between model-based iterative reconstruction and adaptive statistical iterative reconstruction chest computed tomography in cystic fibrosis patients. J Med Imaging Radiat Oncol 2019;63:602-609 https://doi.org/10.1111/1754-9485.12895
  22. Winklehner A, Karlo C, Puippe G, Schmidt B, Flohr T, Goetti R, et al. Raw data-based iterative reconstruction in body CTA: evaluation of radiation dose saving potential. Eur Radiol 2011;21:2521-2526 https://doi.org/10.1007/s00330-011-2227-y
  23. Kuo Y, Lin YY, Lee RC, Lin CJ, Chiou YY, Guo WY. Comparison of image quality from filtered back projection, statistical iterative reconstruction, and model-based iterative reconstruction algorithms in abdominal computed tomography. Medicine (Baltimore) 2016;95:e4456
  24. Gulliksrud K, Stokke C, Martinsen AC. How to measure CT image quality: variations in CT-numbers, uniformity and low contrast resolution for a CT quality assurance phantom. Phys Med 2014;30:521-526 https://doi.org/10.1016/j.ejmp.2014.01.006
  25. Svanholm H, Starklint H, Gundersen HJ, Fabricius J, Barlebo H, Olsen S. Reproducibility of histomorphologic diagnoses with special reference to the kappa statistic. APMIS 1989;97:689-698 https://doi.org/10.1111/j.1699-0463.1989.tb00464.x
  26. Bushberg JT, Seibert A, Boone JM, Leidholdt EM. The essential physics of medical imaging. Philadelphia, PA: Wolters Kluwer/Lippincott Williams & Wilkins, 2012
  27. Rose A. Quantum effects in human vision. Adv Biol Med Phys 1957;5:211-242 https://doi.org/10.1016/B978-1-4832-3111-2.50009-2
  28. Faber J, Fonseca LM. How sample size influences research outcomes. Dental Press J Orthod 2014;19:27-29 https://doi.org/10.1590/2176-9451.19.4.027-029.ebo