Advertisement

Noise reduction and motion elimination in low-dose 4D myocardial computed tomography perfusion (CTP): preliminary clinical evaluation of the ASTRA4D algorithm

  • Steffen Lukas
  • Sarah Feger
  • Matthias Rief
  • Elke Zimmermann
  • Marc DeweyEmail author
Cardiac
  • 17 Downloads

Abstract

Objectives

To propose and evaluate a four-dimensional (4D) algorithm for joint motion elimination and spatiotemporal noise reduction in low-dose dynamic myocardial computed tomography perfusion (CTP).

Methods

Thirty patients with suspected or confirmed coronary artery disease were prospectively included and underwent dynamic contrast-enhanced 320-row CTP. A novel deformable image registration method based on the principal component analysis (PCA) of the ante hoc temporally smoothed voxel-wise time-attenuation curves (ASTRA4D) is presented. Quantitative (standard deviation, signal-to-noise ratio (SNR), temporal variation, volumetric deformation) and qualitative (motion, contrast, contour sharpness [1, poor; 5, excellent]) measures of CTP quality were assessed for the original and motion-compensated sequences (without and with temporal filtering, PCA/ASTRA4D). Following myocardial perfusion deficit detection by two readers, diagnostic accuracy was evaluated using magnetic resonance myocardial perfusion imaging (MR-MPI) as the reference standard in 15 patients.

Results

Registration using ASTRA4D was successful in all 30 patients and resulted in comparison with the benchmark PCA in significantly (p < 0.001) reduced noise over time (− 83%, 178.5 vs 29.9) and spatially (− 34%, 21.4 vs 14.1) as well as improved SNR (+ 47%, 3.6 vs 5.3) and subjective image quality (motion, contrast, contour sharpness [+ 1.0, + 1.0, + 0.5]). ASTRA4D had significantly improved per-segment sensitivity of 91% (58/64) and similar specificity of 96% (429/446) compared with PCA (52%, 33/64; 98%, 435/446; p = 0.011) in the visual detection of perfusion deficits.

Conclusions

The ASTRA4D registration algorithm improved the spatiotemporal noise profile and CTP sequence image quality, resulting in significantly improved sensitivity of 4D CTP in the detection of myocardial ischemia.

Key Points

• ASTRA4D combines local temporal regression and deformable image registration.

• Quantitative and qualitative measures of CTP quality are improved compared to PCA.

• Improved spatiotemporal differentiation of ischemic regions leads to an excellent perfusion deficit concordance of ASTRA4D with MRI.

Keywords

Coronary artery disease Computed tomography myocardial perfusion imaging Temporal averaging Motion artifacts Deformable registration 

Abbreviations

CAD

Coronary artery disease

CT

Computed tomography

CTP

Computed tomography perfusion

HU

Hounsfield unit

IQR

Interquartile range

MPI

Myocardial perfusion imaging

MR

Magnetic resonance

PCA

Principal component analysis

ROI

Region of interest

TAC

Time-attenuation curve

WL

Window level

Notes

Acknowledgements

The abstract for this paper was submitted to and accepted for the European Congress of Radiology in Vienna 2018. The presentation with the title Motion elimination in low-dose 4D myocardial computed tomography perfusion (CTP) using the automated smooth temporal registration for analysis of 4D image data (ASTRA) algorithm (B-0762) was held in the session New CT protocols to assess coronary artery and myocardium (SS 703) on the 1st of March 2018.

Funding

Prof. Dewey has received grant support for this study from the Heisenberg Program of the DFG (DE 1361/14-1).

Compliance with ethical standards

Guarantor

The scientific guarantor of this publication is Prof. Dr. Marc Dewey.

Conflict of interest

The authors of this manuscript declare relationships with the following companies:

Prof. Dewey has received grant support from the Heisenberg Program of the DFG for a professorship (DE 1361/14-1) and the FP7 Program of the European Commission for the randomized multicenter DISCHARGE trial (603266-2, HEALTH-2012.2.4.-2).

Prof. Dewey has received lecture fees from Toshiba Medical Systems, Guerbet, Cardiac MR Academy Berlin, and Bayer (Schering-Berlex).

Prof. Dewey is the editor of the Cardiac Section of European Radiology.

Institutional master research agreements exist with Siemens Medical Solutions, Philips Medical Systems, and Toshiba Medical Systems. The terms of these arrangements are managed by the legal department of Charité – Universitätsmedizin Berlin.

Other authors declared no conflicts of interest.

Statistics and biometry

No complex statistical methods were necessary for this paper.

Informed consent

Written informed consent was obtained from all patients included in this study.

Ethical approval

Institutional Review Board approval was obtained.

Study subjects or cohorts overlap

Study subjects have been previously reported in Feger et al [12].

Methodology

• Retrospective

Supplementary material

330_2018_5899_MOESM1_ESM.docx (37 kb)
ESM 1 (DOCX 36.8 kb)
330_2018_5899_Fig7_ESM.png (36 kb)
ESM2

Quantitative evaluation. Objective image quality performance of the proposed registration method for the myocardium using temporal kernel bandwidths (0-4; 0=PCA, 2=ASTRA4D). ASTRA4D reduced temporal variation by 83% (TV 178.5 vs 29.9) and standard deviation by 34% (SD 21.4 vs 14.1) while it increased signal-to-noise ratio by 47% (SNR 3.6 vs 5.3) compared with the benchmark PCA (p<0.001). In contrast, PCA alone reduced TV by 2% and SD by 9% while it increased SNR by 9% vs the original sequence (p<0.001). (PNG 35.8 kb)

330_2018_5899_MOESM2_ESM.tiff (2.4 mb)
Quantitative evaluation. Objective image quality performance of the proposed registration method for the myocardium using temporal kernel bandwidths (0-4; 0=PCA, 2=ASTRA4D). ASTRA4D reduced temporal variation by 83% (TV 178.5 vs 29.9) and standard deviation by 34% (SD 21.4 vs 14.1) while it increased signal-to-noise ratio by 47% (SNR 3.6 vs 5.3) compared with the benchmark PCA (p<0.001). In contrast, PCA alone reduced TV by 2% and SD by 9% while it increased SNR by 9% vs the original sequence (p<0.001). (TIFF 2.35 mb)
330_2018_5899_MOESM3_ESM.gif (1.6 mb)
Video 1 A representative cardiac long-axis view in the three sequences ORG, PCA and ASTRA. Whereas PCA removes deformation, ASTRA in addition removes noise. (GIF 1609 kb)
330_2018_5899_MOESM4_ESM.gif (1.2 mb)
Video 2 A representative wobbly anatomy in cardiac long-axis view in the three sequences ORG, PCA and ASTRA. Contrary to ASTRA, the application of PCA alone cannot resolve a fluctuating anatomy. (GIF 1253 kb)

References

  1. 1.
    Napp AE, Haase R, Laule M et al (2017) Computed tomography versus invasive coronary angiography: design and methods of the pragmatic randomised multicentre DISCHARGE trial. Eur Radiol 27:2957–2968PubMedGoogle Scholar
  2. 2.
    Bamberg F, Becker A, Schwarz F et al (2011) Detection of hemodynamically significant coronary artery stenosis: incremental diagnostic value of dynamic CT-based myocardial perfusion imaging. Radiology 260:689–698PubMedGoogle Scholar
  3. 3.
    George RT, Jerosch-Herold M, Silva C et al (2007) Quantification of myocardial perfusion using dynamic 64-detector computed tomography. Invest Radiol 42:815–822PubMedGoogle Scholar
  4. 4.
    So A, Wisenberg G, Islam A et al (2012) Non-invasive assessment of functionally relevant coronary artery stenoses with quantitative CT perfusion: preliminary clinical experiences. Eur Radiol 22:39–50PubMedGoogle Scholar
  5. 5.
    Varga-Szemes A, Meinel FG, De Cecco CN, Fuller SR, Bayer RR 2nd, Schoepf UJ (2015) CT myocardial perfusion imaging. AJR Am J Roentgenol 204:487–497Google Scholar
  6. 6.
    de Jong MC, Genders TS, van Geuns RJ, Moelker A, Hunink MG (2012) Diagnostic performance of stress myocardial perfusion imaging for coronary artery disease: a systematic review and meta-analysis. Eur Radiol 22:1881–1895PubMedPubMedCentralGoogle Scholar
  7. 7.
    Rief M, Chen MY, Vavere AL et al (2018) Coronary artery disease: analysis of diagnostic performance of CT perfusion and MR perfusion imaging in comparison with quantitative coronary angiography and SPECT-multicenter prospective trial. Radiology 286:461–470PubMedGoogle Scholar
  8. 8.
    Takx RA, Blomberg BA, El Aidi H et al (2015) Diagnostic accuracy of stress myocardial perfusion imaging compared to invasive coronary angiography with fractional flow reserve meta-analysis. Circ Cardiovasc Imaging 8(1):e002666Google Scholar
  9. 9.
    Williams MC, Mirsadraee S, Dweck MR et al (2017) Computed tomography myocardial perfusion vs (15)O-water positron emission tomography and fractional flow reserve. Eur Radiol 27:1114–1124PubMedGoogle Scholar
  10. 10.
    Williams MC, Newby DE (2016) CT myocardial perfusion imaging: current status and future directions. Clin Radiol 71:739–749PubMedGoogle Scholar
  11. 11.
    Kikuchi Y, Oyama-Manabe N, Naya M et al (2014) Quantification of myocardial blood flow using dynamic 320-row multi-detector CT as compared with (1)(5)O-H(2) O PET. Eur Radiol 24:1547–1556PubMedGoogle Scholar
  12. 12.
    Feger S, Shaban A, Lukas S et al (2017) Temporal averaging for analysis of four-dimensional whole-heart computed tomography perfusion of the myocardium: proof-of-concept study. Int J Cardiovasc Imaging 33:371–382PubMedGoogle Scholar
  13. 13.
    Li Z, Yu L, Leng S et al (2016) A robust noise reduction technique for time resolved CT. Med Phys 43:347PubMedGoogle Scholar
  14. 14.
    Pisana F, Henzler T, Schönberg S, Klotz E, Schmidt B, Kachelrieß M (2017) Noise reduction and functional maps image quality improvement in dynamic CT perfusion using a new k-means clustering guided bilateral filter (KMGB). Med Phys 44:3464–3482PubMedGoogle Scholar
  15. 15.
    Feng Q, Zhou Y, Li X et al (2016) Liver DCE-MRI registration in manifold space based on robust principal component analysis. Sci Rep 6:34461PubMedPubMedCentralGoogle Scholar
  16. 16.
    Hamy V, Dikaios N, Punwani S et al (2014) Respiratory motion correction in dynamic MRI using robust data decomposition registration—application to DCE-MRI. Med Image Anal 18:301–313PubMedGoogle Scholar
  17. 17.
    Huizinga W, Poot DH, Guyader JM et al (2016) PCA-based groupwise image registration for quantitative MRI. Med Image Anal 29:65–78PubMedGoogle Scholar
  18. 18.
    Melbourne A, Atkinson D, White MJ, Collins D, Leach M, Hawkes D (2007) Registration of dynamic contrast-enhanced MRI using a progressive principal component registration (PPCR). Phys Med Biol 52:5147–5156PubMedGoogle Scholar
  19. 19.
    Wollny G, Kellman P, Santos A, Ledesma-Carbayo MJ (2012) Automatic motion compensation of free breathing acquired myocardial perfusion data by using independent component analysis. Med Image Anal 16:1015–1028PubMedPubMedCentralGoogle Scholar
  20. 20.
    Mihai G, Ding Y, Xue H et al (2012) Non-rigid registration and KLT filter to improve SNR and CNR in GRE-EPI myocardial perfusion imaging. J Biomed Sci Eng 5:871–877PubMedPubMedCentralGoogle Scholar
  21. 21.
    Muenzel D, Kabus S, Gramer B et al (2013) Dynamic CT perfusion imaging of the myocardium: a technical note on improvement of image quality. PLoS One 8(10):e75263Google Scholar
  22. 22.
    Feger S, Rief M, Zimmermann E et al (2015) The impact of different levels of adaptive iterative dose reduction 3D on image quality of 320-row coronary CT angiography: a clinical trial. PLoS One 10:e0125943PubMedPubMedCentralGoogle Scholar
  23. 23.
    Techasith T, Cury RC (2011) Stress myocardial CT perfusion: an update and future perspective. JACC Cardiovasc Imaging 4:905–916PubMedGoogle Scholar
  24. 24.
    Cerqueira MD, Weissman NJ, Dilsizian V et al (2002) Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart. A statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Int J Cardiovasc Imaging 18:539–542PubMedGoogle Scholar
  25. 25.
    Bamberg F, Marcus RP, Becker A et al (2014) Dynamic myocardial CT perfusion imaging for evaluation of myocardial ischemia as determined by MR imaging. JACC Cardiovasc Imaging 7:267–277PubMedGoogle Scholar
  26. 26.
    Bischoff B, Bamberg F, Marcus R et al (2013) Optimal timing for first-pass stress CT myocardial perfusion imaging. Int J Cardiovasc Imaging 29:435–442PubMedGoogle Scholar
  27. 27.
    Yang Z, Silver MD (2015) Denoising method and system for preserving clinically significant structures in reconstructed images using adaptively weighted anisotropic diffusion filter. Google PatentsGoogle Scholar
  28. 28.
    So A, Imai Y, Nett B et al (2016) Technical note: evaluation of a 160-mm/256-row CT scanner for whole-heart quantitative myocardial perfusion imaging. Med Phys 43:4821PubMedGoogle Scholar
  29. 29.
    Modgil D, Bindschadler MD, Alessio AM, La Rivière PJ (2017) Variable temporal sampling and tube current modulation for myocardial blood flow estimation from dose-reduced dynamic computed tomography. J Med Imaging (Bellingham) 4:026002Google Scholar

Copyright information

© European Society of Radiology 2019

Authors and Affiliations

  1. 1.Department of RadiologyCharité Medical SchoolBerlinGermany

Personalised recommendations