Objectives
Interscan reproducibility of coronary artery calcium (CAC) scoring can be improved
by using a smaller slice thickness but at the cost of higher image noise. This study
aimed to investigate the feasibility of using densely connected convolutional network
(DenseNet) to reduce the image noise in CAC scans reconstructed with slice thickness
< 3 mm for improving coronary calcification detection in CT.
Methods
Phantom data acquired with QRM and CIRS phantoms were used for model training and
testing, where the DenseNet model adopted in this work was a convolutional neural
network (CNN) designed for super resolution recovery. After phantom study, the proposed
method was evaluated in terms of its ability to improve calcification detection using
patient data. The CNN input images () were CAC scans reconstructed with 0.5-, 1.0- and 1.5-mm slice thickness, while CNN
label images were CAC scans reconstructed with 3-mm slice thickness (). Region of interest (ROI) analysis was carried out on , and CNN output images (). Two-sample t test was used to compare the difference in Hounsfield Unit (HU) values
within ROI between and .
Results
For the calcifications in QRM phantoms, no statistically significant difference was
found when comparing the HU values of 400- and 800-HA calcifications identified on
to those on with slice thickness of 0.5, 1.0 or 1.5 mm. On the other hand, statistically significant
difference was found when comparing the HU values of 200-HA calcifications identified
on to those on with a slice thickness of 0.5 and 1.0 mm. Meanwhile, no statistically significant
difference was found when comparing the HU values of 200-HA calcifications identified
on to those on with a slice thickness of 1.5 mm. As for the rod inserts in CIRS phantoms simulating
9 different tissue types in human body, there was no statistically significant difference
between and with slice thickness of 1.5 mm, and all the p values were larger than 0.10. With regards to patient study, more calcification pixels
were detected on with a slice thickness of 1.5 mm than on , so calcifications were more clear on the denoised images.
Conclusion
According to our results, the CNN-based denoising method could reduce statistical
noise in with a slice thickness of 1.5 mm without causing significant texture change or variation
in HU values. The proposed method could improve cardiovascular risk prediction by
detecting small and soft calcifications that are barely identified on 3-mm slice images
used in conventional CAC scans.
Key words
To read this article in full you will need to make a payment
Purchase one-time access:
Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online accessOne-time access price info
- For academic or personal research use, select 'Academic and Personal'
- For corporate R&D use, select 'Corporate R&D Professionals'
Subscribe:
Subscribe to Academic RadiologyAlready a print subscriber? Claim online access
Already an online subscriber? Sign in
Register: Create an account
Institutional Access: Sign in to ScienceDirect
References
- Assessment of coronary artery disease by cardiac computed tomography: a scientific statement from the American Heart Association Committee on Cardiovascular Imaging and Intervention, Council on Cardiovascular Radiology and Intervention, and Committee on Cardiac Imaging, Council on Clinical Cardiology.Circulation. 2006; 114: 1761-1791
- Coronary calcium score and cardiovascular risk.J Am Coll Cardiol. 2018; 72: 434-447
- Coronary artery calcium scoring: is it time for a change in methodology?.JACC Cardiovasc Imaging. 2017; 10: 923-937
- Coronary artery calcium scoring: current status and review of literature.J Comput Assist Tomogr. 2018; 42: 887-897
- Coronary calcium risk score and cardiovascular risk.Current Vasc Pharmacol. 2021; 19: 280-284
- Coronary artery calcification scoring in low-dose ungated CT screening for lung cancer: interscan agreement.AJR Am J Roentgenol. 2010; 194: 1244-1249
- Coronary artery calcium scoring using ECG-gated multidetector CT: effect of individually optimized image-reconstruction windows on image quality and measurement reproducibility.AJR Am J Roentgenol. 2003; 181: 1093-1100
- Coronary calcium scoring with partial volume correction in anthropomorphic thorax phantom and screening chest CT images.PloS one. 2018; 13e0209318
- Measurement of coronary artery calcium volume using ultra-high-resolution computed tomography: a preliminary phantom and cadaver study.Eur J Radiol Open. 2020; 7100253
- Assessment of isotropic calcium using 0.5-mm reconstructions from 320-row CT data sets identifies more patients with non-zero Agatston score and more subclinical atherosclerosis than standard 3.0-mm coronary artery calcium scan and CT angiography.J Cardiovasc Comput Tomogr. 2014; 8: 58-66
- Identifying small coronary calcification in non-contrast 0.5-mm slice reconstruction to diagnose coronary artery disease in patients with a conventional zero coronary artery calcium score.J Atheroscler Thromb. 2016; 23: 1324-1333
- Accuracy of coronary artery calcium scoring with tube current reduction by 75%, using an adaptive iterative reconstruction algorithm.Br J Radiol. 2018; 9120170678
- Radiation dose modulation techniques in the multidetector CT era: from basics to practice.Radiographics. 28. 2008: 1451-1459
- Quantitative assessment of image noise and streak artifact on CT image: comparison of z-axis automatic tube current modulation technique with fixed tube current technique.Comput Med Imaging Graph. 2009; 33: 353-358
- Innovations in CT dose reduction strategy: application of the adaptive statistical iterative reconstruction algorithm.AJR Am J Roentgenol. 2010; 194: 191-199
- Achievable dose reduction using iterative reconstruction for chest computed tomography: a systematic review.Eur J Radiol. 2015; 84: 2307-2313
- The use of deep learning towards dose optimization in low-dose computed tomography: a scoping review.Radiography. 2022; 28: 208-214
- The potential for reduced radiation dose from deep learning-based CT image reconstruction: A comparison with filtered back projection and hybrid iterative reconstruction using a phantom.Medicine. 2021; 100: 1-9
- A review on deep learning approaches for low-dose computed tomography restoration.Complex Intell Systems. 2021; : 1-33
- Review of deep learning algorithms and architectures.IEEE Access. 2019; 7: 53040-53065
- Two stage residual CNN for texture denoising and structure enhancement on low dose CT image.Comput Methods Programs Biomed. 2020; 184105115
- Low-dose CT lung images denoising based on multiscale parallel convolution neural network.Visual Computer. 2021; 37: 1-13
- Ultra-low-dose chest CT imaging of COVID-19 patients using a deep residual neural network.Eur Radiol. 2021; 31: 1420-1431
- Image super-resolution using dense skip connections.in: Paper presented at: Venice, Italy: 2017 IEEE International Conference on Computer Vision (ICCV); 22-29 Oct. 2017. 2017
- Delving deep into rectifiers: surpassing human-level performance on Imagenet classification.in: Santiago, Chile: IEEE International Conference on Computer Vision (ICCV 2015). 2015: 1502
- A method for stochastic optimization.in: Banff, Canada: International Conference on Learning Representations. 2014
- Caffe: convolutional architecture for fast feature embedding.in: Orlando, USA: MM 2014 - Proceedings of the 2014 ACM Conference on Multimedia. 2014
Article info
Publication history
Published online: November 15, 2022
Accepted:
October 18,
2022
Received in revised form:
October 5,
2022
Received:
September 15,
2022
Publication stage
In Press Corrected ProofIdentification
Copyright
© 2022 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.