Abstract
The emergence of 7T MRI scanners offers groundbreaking potential in medical imaging, providing ultra-high resolution compared to standard 3T MRI machines. However, widespread adoption faces challenges due to high costs and safety concerns. Our research addresses this by enhancing 3T MRI image quality to rival 7T scanners, particularly in cases where skull stripping is difficult. We propose a dense U-Net algorithm, surpassing previous deep learning methods, achieving an average PSNR of 23.8638 ± 2.3706 and SSIM of 0.7525 ± 0.0646. Validation on a paired MRI dataset of 3 patients scanned in both 3T and 7T machines showed promising results. Neuroradiologists ranked our tool highest in overall image quality, anatomical structure, and diagnostic confidence compared to two others on the Likert scale. Our proposed method could facilitate broader access to advanced imaging capabilities in clinical settings.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ahmad, W., Ali, H., Shah, Z., et al.: A new generative adversarial network for medical images super resolution. Sci. Rep. 12, 9533 (2022). https://doi.org/10.1038/s41598-022-13658-4
Gao, Y., Li, H., Dong, J., Feng, G.:A deep convolutional network for medical image super-resolution. In: 2017 Chinese Automation Congress (CAC), Jinan, China, pp. 5310–5315 (2017).https://doi.org/10.1109/CAC.2017.8243724
Mathivanan, S.K., Sonaimuthu, S., Murugesan, S., et al.: Employing deep learning and transfer learning for accurate brain tumor detection. Sci. Rep. 14, 7232 (2024). https://doi.org/10.1038/s41598-024-57970-7
Kalluvila, A.: Super-resolution of brain MRI via U-Net architecture. Int. J. Adv. Comput. Sci. Appl. (IJACSA), 14(5) (2023). https://doi.org/10.14569/IJACSA.2023.0140503
Isaacs, B.R., et al.: 3 versus 7 Tesla magnetic resonance imaging for parcellations of subcortical brain structures in clinical settings. PLoS ONE 15(11), e0236208 (2020). https://doi.org/10.1371/journal.pone.0236208.PMID:33232325;PMCID:PMC7685480
Chen, X., Qu, L., Xie, Y., et al.: A paired dataset of T1- and T2-weighted MRI at 3 Tesla and 7 Tesla. Sci Data 10, 489 (2023). https://doi.org/10.1038/s41597-023-02400-y
GeeksforGeeks. U-Net Architecture explained (2023). https://www.geeksforgeeks.org/u-net-architecture-explained/
Acknowledgements
The authors would like to thank Dr. Benjamin Pulli (B.J.) and Dr. Jeremy Heit (J.H.) for serving as the blind radiologist reviewers for this study.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Ethics declarations
Disclosure of Interests
The authors have no conflicts of interest.
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Kalluvila, A., Rosen, M.S. (2024). 3T to 7T Whole Brain + Skull MRI Translation with Densely Engineered U-Net Network. In: Finkelstein, J., Moskovitch, R., Parimbelli, E. (eds) Artificial Intelligence in Medicine. AIME 2024. Lecture Notes in Computer Science(), vol 14845. Springer, Cham. https://doi.org/10.1007/978-3-031-66535-6_1
Download citation
DOI: https://doi.org/10.1007/978-3-031-66535-6_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-66534-9
Online ISBN: 978-3-031-66535-6
eBook Packages: Computer ScienceComputer Science (R0)