nach oben

European Journal of Nuclear Medicine and Molecular Imaging

Erschienen in:

12.03.2022 | Original Article

Deep learning–based attenuation correction for whole-body PET — a multi-tracer study with ¹⁸F-FDG, ⁶⁸ Ga-DOTATATE, and ¹⁸F-Fluciclovine

verfasst von: Takuya Toyonaga, Dan Shao, Luyao Shi, Jiazhen Zhang, Enette Mae Revilla, David Menard, Joseph Ankrah, Kenji Hirata, Ming-Kai Chen, John A. Onofrey, Yihuan Lu

Erschienen in: European Journal of Nuclear Medicine and Molecular Imaging | Ausgabe 9/2022

Einloggen, um Zugang zu erhalten

Abstract

A novel deep learning (DL)-based attenuation correction (AC) framework was applied to clinical whole-body oncology studies using ¹⁸F-FDG, ⁶⁸ Ga-DOTATATE, and ¹⁸F-Fluciclovine. The framework used activity (λ-MLAA) and attenuation (µ-MLAA) maps estimated by the maximum likelihood reconstruction of activity and attenuation (MLAA) algorithm as inputs to a modified U-net neural network with a novel imaging physics-based loss function to learn a CT-derived attenuation map (µ-CT).

Methods

Clinical whole-body PET/CT datasets of ¹⁸F-FDG (N = 113), ⁶⁸ Ga-DOTATATE (N = 76), and ¹⁸F-Fluciclovine (N = 90) were used to train and test tracer-specific neural networks. For each tracer, forty subjects were used to train the neural network to predict attenuation maps (µ-DL). µ-DL and µ-MLAA were compared to the gold-standard µ-CT. PET images reconstructed using the OSEM algorithm with µ-DL (OSEM_DL) and µ-MLAA (OSEM_MLAA) were compared to the CT-based reconstruction (OSEM_CT). Tumor regions of interest were segmented by two radiologists and tumor SUV and volume measures were reported, as well as evaluation using conventional image analysis metrics.

Results

µ-DL yielded high resolution and fine detail recovery of the attenuation map, which was superior in quality as compared to µ-MLAA in all metrics for all tracers. Using OSEM_CT as the gold-standard, OSEM_DL provided more accurate tumor quantification than OSEM_MLAA for all three tracers, e.g., error in SUV_max for OSEM_MLAA vs. OSEM_DL: − 3.6 ± 4.4% vs. − 1.7 ± 4.5% for ¹⁸F-FDG (N = 152), − 4.3 ± 5.1% vs. 0.4 ± 2.8% for ⁶⁸ Ga-DOTATATE (N = 70), and − 7.3 ± 2.9% vs. − 2.8 ± 2.3% for ¹⁸F-Fluciclovine (N = 44). OSEM_DL also yielded more accurate tumor volume measures than OSEM_MLAA, i.e., − 8.4 ± 14.5% (OSEM_MLAA) vs. − 3.0 ± 15.0% for ¹⁸F-FDG, − 14.1 ± 19.7% vs. 1.8 ± 11.6% for ⁶⁸ Ga-DOTATATE, and − 15.9 ± 9.1% vs. − 6.4 ± 6.4% for ¹⁸F-Fluciclovine.

Conclusions

The proposed framework provides accurate and robust attenuation correction for whole-body ¹⁸F-FDG, ⁶⁸ Ga-DOTATATE and ¹⁸F-Fluciclovine in tumor SUV measures as well as tumor volume estimation. The proposed method provides clinically equivalent quality as compared to CT in attenuation correction for the three tracers.

Nur mit Berechtigung zugänglich

Barrett JF, Keat N. Artifacts in CT: recognition and avoidance. Radiographics. 2004;24:1679–91. https://doi.org/10.1148/rg.246045065.CrossRefPubMed

Ladefoged CN, Law I, Anazodo U, St Lawrence K, Izquierdo-Garcia D, Catana C, et al. A multi-centre evaluation of eleven clinically feasible brain PET/MRI attenuation correction techniques using a large cohort of patients. Neuroimage. 2017;147:346–59. https://doi.org/10.1016/j.neuroimage.2016.12.010.CrossRefPubMed

Chen Y, An H. Attenuation correction of PET/MR imaging. Magn Reson Imaging Clin N Am. 2017;25:245–55. https://doi.org/10.1016/j.mric.2016.12.001.CrossRefPubMedPubMedCentral

Rezaei A, Defrise M, Bal G, Michel C, Conti M, Watson C, et al. Simultaneous reconstruction of activity and attenuation in time-of-flight PET. IEEE Trans Med Imaging. 2012;31:2224–33. https://doi.org/10.1109/TMI.2012.2212719.CrossRefPubMed

Rezaei A, Deroose CM, Vahle T, Boada F, Nuyts J. Joint reconstruction of activity and attenuation in time-of-flight pet: a quantitative analysis. J Nucl Med. 2018;59:1630. https://doi.org/10.2967/jnumed.117.204156.CrossRefPubMedPubMedCentral

Hwang D, Kang SK, Kim KY, Seo S, Paeng JC, Lee DS, et al. Generation of PET attenuation map for whole-body time-of-flight (18)F-FDG PET/MRI using a deep neural network trained with simultaneously reconstructed activity and attenuation maps. J Nucl Med. 2019;60:1183–9. https://doi.org/10.2967/jnumed.118.219493.CrossRefPubMedPubMedCentral

Shi L, Onofrey J, Revilla EM, Toyonaga T, Menard D, Ankrah J, et al. A novel loss function incorporating imaging acquisition physics for PET attenuation map generation using deep learning. Med Image Comput Comput Assist Interv. 2019;11767:723–31. https://doi.org/10.1007/978-3-030-32251-9_79.CrossRef

Maurer T, Eiber M, Schwaiger M, Gschwend JE. Current use of PSMA-PET in prostate cancer management. Nat Rev Urol. 2016;13:226–35. https://doi.org/10.1038/nrurol.2016.26.CrossRefPubMed

Calais J, Ceci F, Eiber M, Hope TA, Hofman MS, Rischpler C, et al. (18)F-fluciclovine PET-CT and (68)Ga-PSMA-11 PET-CT in patients with early biochemical recurrence after prostatectomy: a prospective, single-centre, single-arm, comparative imaging trial. Lancet Oncol. 2019;20:1286–94. https://doi.org/10.1016/S1470-2045(19)30415-2.CrossRefPubMedPubMedCentral

10.

Poeppel TD, Binse I, Petersenn S, Lahner H, Schott M, Antoch G, et al. 68Ga-DOTATOC versus 68Ga-DOTATATE PET/CT in functional imaging of neuroendocrine tumors. J Nucl Med. 2011;52:1864–70. https://doi.org/10.2967/jnumed.111.091165.CrossRefPubMed

11.

Panin VY, Aykac M, Casey ME. Simultaneous reconstruction of emission activity and attenuation coefficient distribution from TOF data, acquired with external transmission source. Phys Med Biol. 2013;58:3649–69. https://doi.org/10.1088/0031-9155/58/11/3649.CrossRefPubMed

12.

Onofrey JA, Casetti-Dinescu DI, Lauritzen AD, Sarkar S, Venkataraman R, Fan RE, et al. Generalizable multi-site training and testing of deep neural networks using image normalization. Biomedical Imaging (ISBI), 2019 IEEE 16th International Symposium on; 2019. p. pp. 1–4.

13.

Ronneberger O, Fischer P, Brox T. U-Net: Convolutional networks for biomedical image segmentation. Cham: Springer International Publishing; 2015. p. 234–41.

14.

Shi L, Onofrey JA, Revilla EM, Toyonaga T, Menard D, Ankrah J, et al. A novel loss function incorporating imaging acquisition physics for PET attenuation map generation using deep learning. Cham: Springer International Publishing; 2019. p. 723–31.

15.

Nie D, Trullo R, Lian J, Wang L, Petitjean C, Ruan S, et al. Medical image synthesis with deep convolutional adversarial networks. Ieee T Bio-Med Eng. 2018;65:2720–30. https://doi.org/10.1109/Tbme.2018.2814538.CrossRef

16.

Kingma DP, Ba J. Adam: a method for stochastic optimization. 2014. p. arXiv:1412.6980.

17.

Onofrey JA, Casetti-Dinescu DI, Lauritzen AD, Sarkar S, Venkataraman R, Fan RE, et al. Generalizable multi-site training and testing of deep neural networks using image normalization. Proc IEEE Int Symp Biomed Imaging. 2019;2019:348–51. https://doi.org/10.1109/isbi.2019.8759295.CrossRefPubMedPubMedCentral

18.

Hirata K, Furuya S, Huang SC, Manabe O, Magota K, Kobayashi K, et al. A semi-automated method to separate tumor from physiological uptakes on FDG PET-CT for efficient generation of training data targeting deep learning. J Nucl Med. 2019;60:supplement 1213.CrossRef

19.

Bradshaw TJ, Zhao G, Jang H, Liu F, McMillan AB. Feasibility of deep learning-based PET/MR attenuation correction in the pelvis using only diagnostic MR images. Tomography. 2018;4:138–47. https://doi.org/10.18383/j.tom.2018.00016.CrossRefPubMedPubMedCentral

20.

Arabi H, Zeng G, Zheng G, Zaidi H. Novel adversarial semantic structure deep learning for MRI-guided attenuation correction in brain PET/MRI. Eur J Nucl Med Mol Imaging. 2019;46:2746–59. https://doi.org/10.1007/s00259-019-04380-x.CrossRefPubMed

21.

Liu F, Jang H, Kijowski R, Zhao G, Bradshaw T, McMillan AB. A deep learning approach for (18)F-FDG PET attenuation correction. EJNMMI Phys. 2018;5:24. https://doi.org/10.1186/s40658-018-0225-8.CrossRefPubMedPubMedCentral

22.

Dong X, Wang T, Lei Y, Higgins K, Liu T, Curran WJ, et al. Synthetic CT generation from non-attenuation corrected PET images for whole-body PET imaging. Phys Med Biol. 2019;64:215016. https://doi.org/10.1088/1361-6560/ab4eb7.CrossRefPubMedPubMedCentral

23.

Shiri I, Ghafarian P, Geramifar P, Leung KH, Ghelichoghli M, Oveisi M, et al. Direct attenuation correction of brain PET images using only emission data via a deep convolutional encoder-decoder (Deep-DAC). Eur Radiol. 2019;29:6867–79. https://doi.org/10.1007/s00330-019-06229-1.CrossRefPubMed

24.

Hashimoto F, Ito M, Ote K, Isobe T, Okada H, Ouchi Y. Deep learning-based attenuation correction for brain PET with various radiotracers. Ann Nucl Med. 2021. https://doi.org/10.1007/s12149-021-01611-w.CrossRefPubMedPubMedCentral

25.

Lee JS. A review of deep-learning-based approaches for attenuation correction in positron emission tomography. IEEE Transactions on Radiation and Plasma Medical Sciences. 2021;5:160–84. https://doi.org/10.1109/TRPMS.2020.3009269.CrossRef

26.

Gong K, Guan J, Kim K, Zhang X, Yang J, Seo Y, et al. Iterative PET Image reconstruction using convolutional neural network representation. Ieee T Med Imaging. 2019;38:675–85. https://doi.org/10.1109/TMI.2018.2869871.CrossRef

27.

Lu W, Onofrey JA, Lu Y, Shi L, Ma T, Liu Y, et al. An investigation of quantitative accuracy for deep learning based denoising in oncological PET. Phys Med Biol. 2019;64:165019. https://doi.org/10.1088/1361-6560/ab3242.CrossRefPubMed

28.

Li Y, Jiang L, Wang H, Cai H, Xiang Y, Li L. Effective radiation dose of 18f-Fdg Pet/Ct: how much does diagnostic Ct contribute? Radiat Prot Dosimetry. 2019;187:183–90. https://doi.org/10.1093/rpd/ncz153.CrossRefPubMed

29.

Lu Y, Gallezot JD, Naganawa M, Ren S, Fontaine K, Wu J, et al. Data-driven voluntary body motion detection and non-rigid event-by-event correction for static and dynamic PET. Phys Med Biol. 2019;64:065002. https://doi.org/10.1088/1361-6560/ab02c2.CrossRefPubMed

30.

Lu Y, Fontaine K, Mulnix T, Onofrey JA, Ren S, Panin V, et al. Respiratory motion compensation for PET/CT with motion information derived from matched attenuation-corrected gated PET data. J Nucl Med. 2018;59:1480–6. https://doi.org/10.2967/jnumed.117.203000.CrossRefPubMedPubMedCentral

31.

Teimoorisichani M, Sari H, Panin V, Bharkhada D, Rominger A, Conti M. Using LSO background radiation for CT-less attenuation correction of PET data in long axial FOV PET scanners. Journal of Nuclear Medicine. 2021;62:1530-.

32.

Rothfuss H, Panin V, Moor A, Young J, Hong I, Michel C, et al. LSO background radiation as a transmission source using time of flight. Phys Med Biol. 2014;59:5483–500. https://doi.org/10.1088/0031-9155/59/18/5483.CrossRefPubMed

Titel: Deep learning–based attenuation correction for whole-body PET — a multi-tracer study with 18F-FDG, 68 Ga-DOTATATE, and 18F-Fluciclovine
verfasst von: Takuya Toyonaga
Dan Shao
Luyao Shi
Jiazhen Zhang
Enette Mae Revilla
David Menard
Joseph Ankrah
Kenji Hirata
Ming-Kai Chen
John A. Onofrey
Yihuan Lu
Publikationsdatum: 12.03.2022
Verlag: Springer Berlin Heidelberg
Erschienen in: European Journal of Nuclear Medicine and Molecular Imaging / Ausgabe 9/2022
Print ISSN: 1619-7070
Elektronische ISSN: 1619-7089
DOI: https://doi.org/10.1007/s00259-022-05748-2

Live-Webinar: Aktuelle Leitlinien bei Herz-Kreislauf-Erkrankungen

Springer Medizin

Deep learning–based attenuation correction for whole-body PET — a multi-tracer study with ¹⁸F-FDG, ⁶⁸ Ga-DOTATATE, and ¹⁸F-Fluciclovine

Abstract

Methods

Results

Conclusions

Live-Webinar: Aktuelle Leitlinien bei Herz-Kreislauf-Erkrankungen

Springer Medizin

Abstract

Methods

Results

Conclusions

Bitte loggen Sie sich ein, um Zugang zu diesem Inhalt zu erhalten

Weitere Artikel der Ausgabe 9/2022

Matched-pair analysis of [177Lu]Lu-PSMA I&T and [177Lu]Lu-PSMA-617 in patients with metastatic castration-resistant prostate cancer

Decoupling of regional neural activity and inter-regional functional connectivity in Alzheimer’s disease: a simultaneous PET/MR study

Correction to: Effects of chronic voluntary alcohol consumption on PDE10A availability: a longitudinal behavioral and [18F]JNJ42259152 PET study in rats

Molecular fluorophores for in vivo bioimaging in the second near-infrared window

Unified spatial normalization method of brain PET images using adaptive probabilistic brain atlas

Abbreviated scan protocols to capture 18F-FDG kinetics for long axial FOV PET scanners