Background

PET MRI plays an important role in the diagnosis and surveillance of neoplastic diseases. PET provides a unique tool for visualising relevant biological processes for patient diagnosis, staging, progression and treatment outcome. In oncology, PET MRI data used in combination with Magnetic Resonane Imaging (MRI) can significantly improve cancer diagnostic accuracy and tumour delineation for radiotherapy treatment planning. This is done by providing vital functional information not otherwise available. In PET interpretation one of the confounding factors is noise in the images, known as background artefact. PET images generally show higher noise levels than other instruments such as Magnetic Resonance Imaging (MRI).
The problem of high noise in PET MRI images is most evident in scenarios where the dose of the radiotracer is reduced to decrease radiation exposure to the patient or scan time. This is desirable in general during all diagnostic procedures not only from view of production but also for dose patient care.
In this hybrid imaging, the PET scan time is set by the operator based on the duration of the MRI sequences. Therefore the optimization of the two exams must be parallel

Objective

The objective is to examine some Artificial Intelligence techniques already described in other studies and evaluate their implementability in PET MRI image reconstruction in order to increase the signal-to-noise ratio and image definition allowing shorter scan times or lower radiopharmaceutical dose administration to the patient.

Materials and Methods

I classified studies for PET image generation with deep learning into three themes: (1) retrieval of complete PET data from noisy data by denoising with deep learning with the aim of reducing scan time, (2) reconstruction of PET images with low count statistics due to reduced radiopharmaceutical doses (3) evaluating the impact of artificial intelligence-based denoising on PET radiomics.
In parallel, I evaluated the image quality and diagnostic performance of MRI with reduced scanning times using denoising approaches with deep learning (DL) based reconstructions in the literature.

Conclusions

This paper describes deep learning techniques that have made significant advances in comprehensive data retrieval, demonstrating that it is feasible to consider implementing tools for reconstructing images from examinations with low scan times or radiopharmaceutical dose and using these images in radiomics studies.

Acknowledgements

No acknowledgement found.

References

1. Shaikh, S. PET-CT Imaging and Applications. In Computed-Tomography (CT) Scan; R. Gharieb, R., Ed.; IntechOpen, 2022 ISBN 978-1-80355-117-3. 2. Gavriilidis, P.; Koole, M.; Annunziata, S.; Mottaghy, F.M.; Wierts, R. Positron Range Corrections and Denoising Techniques for Gallium-68 PET Imaging: A Literature Review. Diagnostics 2022, 12, doi:10.3390/diagnostics12102335. 3. Vriens, D.; Visser, E.P.; de Geus-Oei, L.-F.; Oyen, W.J.G. Methodological Considerations in Quantification of Oncological FDG PET Studies. Eur J Nucl Med Mol Imaging 2010, 37, 1408–1425, doi:10.1007/s00259-009-1306-7. 4. Chang, T.; Chang, G.; Kohlmyer, S.; Clark, J.W.; Rohren, E.; Mawlawi, O.R. Effects of Injected Dose, BMI and Scanner Type on NECR and Image Noise in PET Imaging. Phys. Med. Biol. 2011, 56, 5275–5285, doi:10.1088/0031-9155/56/16/013. 5. Yang, X.; Peng, H. The Use of Noise Equivalent Count Rate and the NEMA Phantom for PET Image Quality Evaluation. Physica Medica 2015, 31, 179–184, doi:10.1016/j.ejmp.2015.01.003. 6. Weyts, K.; Lasnon, C.; Ciappuccini, R.; Lequesne, J.; Corroyer-Dulmont, A.; Quak, E.; Clarisse, B.; Roussel, L.; Bardet, S.; Jaudet, C. Artificial Intelligence-Based PET Denoising Could Allow a Two-Fold Reduction in [18F]FDG PET Acquisition Time in Digital PET/CT. Eur J Nucl Med Mol Imaging 2022, 49, 3750–3760, doi:10.1007/s00259-022-05800-1. 7. Schaefferkoetter, J.; Yan, J.; Ortega, C.; Sertic, A.; Lechtman, E.; Eshet, Y.; Metser, U.; Veit-Haibach, P. Convolutional Neural Networks for Improving Image Quality with Noisy PET Data. EJNMMI Res 2020, 10, 105, doi:10.1186/s13550-020-00695-1. 8. Jaudet, C.; Weyts, K.; Lechervy, A.; Batalla, A.; Bardet, S.; Corroyer-Dulmont, A. The Impact of Artificial Intelligence CNN Based Denoising on FDG PET Radiomics. Front. Oncol. 2021, 11, 692973, doi:10.3389/fonc.2021.692973. 9. Katsari, K.; Penna, D.; Arena, V.; Polverari, G.; Ianniello, A.; Italiano, D.; Milani, R.; Roncacci, A.; Illing, R.O.; Pelosi, E. Artificial Intelligence for Reduced Dose 18F-FDG PET Examinations: A Real-World Deployment through a Standardized Framework and Business Case Assessment. EJNMMI Phys 2021, 8, 25, doi:10.1186/s40658-021-00374-7. 10. Bonardel, G.; Dupont, A.; Decazes, P.; Queneau, M.; Modzelewski, R.; Coulot, J.; Le Calvez, N.; Hapdey, S. Clinical and Phantom Validation of a Deep Learning Based Denoising Algorithm for F-18-FDG PET Images from Lower Detection Counting in Comparison with the Standard Acquisition. EJNMMI Phys 2022, 9, 36, doi:10.1186/s40658-022-00465-z. 11. Gottesman, O.; Johansson, F.; Komorowski, M.; Faisal, A.; Sontag, D.; Doshi-Velez, F.; Celi, L.A. Guidelines for Reinforcement Learning in Healthcare. Nat Med 2019, 25, 16–18, doi:10.1038/s41591-018-0310-5. 12. Chen, K.T.; Gong, E.; de Carvalho Macruz, F.B.; Xu, J.; Boumis, A.; Khalighi, M.; Poston, K.L.; Sha, S.J.; Greicius, M.D.; Mormino, E.; et al. Ultra–Low-Dose 18F-Florbetaben Amyloid PET Imaging Using Deep Learning with Multi-Contrast MRI Inputs. Radiology 2019, 290, 649–656, doi:10.1148/radiol.2018180940. 13. SubtlePET - AI Software for Improving Your PET Scanners. Subtle Medical. 14. Su, R.; Zhang, D.; Liu, J.; Cheng, C. MSU-Net: Multi-Scale U-Net for 2D Medical Image Segmentation. Front. Genet. 2021, 12, 639930, doi:10.3389/fgene.2021.639930. 15. Amirkoushyar Ziabari, Dong Hye Ye †; Somesh Srivastava‡; Ken D. Sauer ⊕; Jean-Baptiste Thibault ‡; Charles A. Bouman 2.5D DEEP LEARNING FOR CT IMAGE RECONSTRUCTION USING A MULTI-GPU IMPLEMENTATION. 16. Chaudhari, A.S.; Mittra, E.; Davidzon, G.A.; Gulaka, P.; Gandhi, H.; Brown, A.; Zhang, T.; Srinivas, S.; Gong, E.; Zaharchuk, G.; et al. Low-Count Whole-Body PET with Deep Learning in a Multicenter and Externally Validated Study. npj Digit. Med. 2021, 4, 127, doi:10.1038/s41746-021-00497-2. 17. Gianoli, C.; Riboldi, M.; Kurz, C.; De Bernardi, E.; Bauer, J.; Fontana, G.; Ciocca, M.; Parodi, K.; Baroni, G. PET-CT Scanner Characterization for PET Raw Data Use in Biomedical Research. Computerized Medical Imaging and Graphics 2014, 38, 358–368, doi:10.1016/j.compmedimag.2014.03.008. 18. Kim, H.S.; Cho, S.-G.; Kim, J.H.; Kwon, S.Y.; Lee, B.-I.; Bom, H.-S. Effect of Post-Reconstruction Gaussian Filtering on Image Quality and Myocardial Blood Flow Measurement with N-13 Ammonia PET. Asia Ocean J Nucl Med Biol 2014, 2, 104–110. 19. Gong, K.; Kim, K.; Cui, J.; Wu, D.; Li, Q. The Evolution of Image Reconstruction in PET. PET Clinics 2021, 16, 533–542, doi:10.1016/j.cpet.2021.06.004. 20. Iriarte, A.; Marabini, R.; Matej, S.; Sorzano, C.O.S.; Lewitt, R.M. System Models for PET Statistical Iterative Reconstruction: A Review. Computerized Medical Imaging and Graphics 2016, 48, 30–48, doi:10.1016/j.compmedimag.2015.12.003. 21. Kaalep, A.; Sera, T.; Rijnsdorp, S.; Yaqub, M.; Talsma, A.; Lodge, M.A.; Boellaard, R. Feasibility of State of the Art PET/CT Systems Performance Harmonisation. Eur J Nucl Med Mol Imaging 2018, 45, 1344–1361, doi:10.1007/s00259-018-3977-4. 22. Aide, N.; Lasnon, C.; Veit-Haibach, P.; Sera, T.; Sattler, B.; Boellaard, R. EANM/EARL Harmonization Strategies in PET Quantification: From Daily Practice to Multicentre Oncological Studies. Eur J Nucl Med Mol Imaging 2017, 44, 17–31, doi:10.1007/s00259-017-3740-2.