Direct attenuation correction of brain PET images using only emission data via a deep convolutional encoder-decoder (Deep-DAC)
To obtain attenuation-corrected PET images directly from non-attenuation-corrected images using a convolutional encoder-decoder network.
Brain PET images from 129 patients were evaluated. The network was designed to map non-attenuation-corrected (NAC) images to pixel-wise continuously valued measured attenuation-corrected (MAC) PET images via an encoder-decoder architecture. Image quality was evaluated using various evaluation metrics. Image quantification was assessed for 19 radiomic features in 83 brain regions as delineated using the Hammersmith atlas (n30r83). Reliability of measurements was determined using pixel-wise relative errors (RE; %) for radiomic feature values in reference MAC PET images.
Peak signal-to-noise ratio (PSNR) and structural similarity index metric (SSIM) values were 39.2 ± 3.65 and 0.989 ± 0.006 for the external validation set, respectively. RE (%) of SUVmean was − 0.10 ± 2.14 for all regions, and only 3 of 83 regions depicted significant differences. However, the mean RE (%) of this region was 0.02 (range, − 0.83 to 1.18). SUVmax had mean RE (%) of − 3.87 ± 2.84 for all brain regions, and 17 regions in the brain depicted significant differences with respect to MAC images with a mean RE of − 3.99 ± 2.11 (range, − 8.46 to 0.76). Homogeneity amongst Haralick-based radiomic features had the highest number (20) of regions with significant differences with a mean RE (%) of 7.22 ± 2.99.
Direct AC of PET images using deep convolutional encoder-decoder networks is a promising technique for brain PET images. The proposed deep learning method shows significant potential for emission-based AC in PET images with applications in PET/MRI and dedicated brain PET scanners.
• We demonstrate direct emission-based attenuation correction of PET images without using anatomical information.
• We performed radiomics analysis of 83 brain regions to show robustness of direct attenuation correction of PET images.
• Deep learning methods have significant promise for emission-based attenuation correction in PET images with potential applications in PET/MRI and dedicated brain PET scanners.
KeywordsPositron emission tomography Brain imaging Artificial intelligence Deep learning Radiomics
Conditional generative adversarial networks
Convolutional neural network
Deep direct attenuation correction
Field of view
Gray-level co-occurrence matrix
Gray-level run length matrix
Gray-level size zone matrix
Graphics processing unit
Measured attenuation corrected
Mean absolute error
Maximum likelihood reconstruction of activity and attenuation
Magnetic resonance imaging
Mean squared error
Ordered subset expectation maximization
Positron emission tomography
Peak signal-to-noise ratio
Restricted Boltzmann machine
Rectified linear unit
Radiomic feature values
Root mean squared error
Structural similarity index metrics
Standard uptake value
Size zone emphasis
Total lesion glycolysis
Time of flight
Ultra-short echo time
Volumes of interest
Zero echo time
This study has received funding the Tehran University of Medical Sciences under grant number 97-01-30-38001.
Compliance with ethical standards
The scientific guarantor of this publication is Mohammad Reza Ay, PhD, Professor of Medical Physics.
Conflict of interest
The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.
Statistics and biometry
One of the authors has significant statistical expertise. And no complex statistical methods were necessary for this paper.
Written informed consent was waived by the Institutional Review Board.
Institutional Review Board approval was obtained.
• Performed at one institution
- 11.Delso G, Nuyts J (2018) PET/MRI: attenuation correction. In: Iagaru A, Hope T, Veit-Haibach P (eds) PET/MRI in Oncology. Springer, Cham, pp 53–75Google Scholar
- 13.Khateri P, Saligheh Rad H, Jafari AH et al (2015) Generation of a four-class attenuation map for MRI-based attenuation correction of PET data in the head area using a novel combination of STE/Dixon-MRI and FCM clustering. Mol Imaging Biol 17:884–892Google Scholar
- 17.Khateri P, Saligheh Rad H, Fathi A, Ay MR (2013) Generation of attenuation map for MR-based attenuation correction of PET data in the head area employing 3D short echo time MR imaging. Nucl Instrum Meth A 702:133–136Google Scholar
- 18.Kazerooni AF, A’arabi MH, Ay M, Saligheh Rad H (2015) Generation of MR-based attenuation correction map of PET images in the brain employing joint segmentation of skull and soft-tissue from single short-TE MR imaging modality. In: Gao F, Shi K, Li S (eds) Computational Methods for Molecular Imaging. Lecture Notes in Computational Vision and Biomechanics, vol 22. Springer, Cham, pp 139–147Google Scholar
- 19.Hope T, Tosun D, Khalighi MM et al (2017) Improvement in quantitative amyloid imaging using ZTE-based attenuation correction in PET/MRI. J Nucl Med 58:645Google Scholar
- 20.Shandiz MS, Saligheh Rad H, Ghafarian P, Yaghoubi K, Ay MR (2018) Capturing bone signal in MRI of pelvis, as a large FOV region, using TWIST sequence and generating a 5-class attenuation map for prostate PET/MRI imaging. Mol Imaging 17:1536012118789314Google Scholar
- 26.Hemmati H, Kamali-Asl A, Ghafarian P, Ay MR (2018) Reconstruction/segmentation of attenuation map in TOF-PET based on mixture models. Ann Nucl Med 1–11, 32(7):474–484Google Scholar
- 28.Yang Q, Li N, Zhao Z, Fan X, Chang EI, Xu Y (2018) MRI image-to-image translation for cross-modality image registration and segmentation. arXiv:180106940Google Scholar
- 30.Ben-Cohen A, Klang E, Raskin SP et al (2018) Cross-modality synthesis from CT to PET using FCN and GAN networks for improved automated lesion detection. arXiv:180207846Google Scholar
- 31.Jin C-B, Jung W, Joo S et al (2018) Deep CT to MR synthesis using paired and unpaired data. arXiv:180510790Google Scholar
- 33.Gong K, Yang J, Kim K, El Fakhri G, Seo Y, Li Q (2018) Attenuation correction for brain PET imaging using deep neural network based on Dixon and ZTE MR images. Phys Med Biol 63(12):125011Google Scholar
- 34.Spuhler KD, Gardus J, Gao Y, DeLorenzo C, Parsey R, Huang C (2018) Synthesis of patient-specific transmission image for PET attenuation correction for PET/MR imaging of the brain using a convolutional neural network. J Nucl Med 60(4):555–560Google Scholar
- 35.Lu Y, Fontaine K, Germino M et al (2018) Investigation of sub-centimeter lung nodule quantification for low-dose PET. IEEE TRPMS 2:41–50Google Scholar
- 36.Abadi M, Barham P, Chen J et al (2016) TensorFlow: a system for large-scale machine learning. In 12th USENIX Symposium on Operating Systems Design and Implementation, pp 265–283Google Scholar