Display options
Share it on

J Imaging. 2021 Dec 03;7(12). doi: 10.3390/jimaging7120262.

Optical to Planar X-ray Mouse Image Mapping in Preclinical Nuclear Medicine Using Conditional Adversarial Networks.

Journal of imaging

Eleftherios Fysikopoulos, Maritina Rouchota, Vasilis Eleftheriadis, Christina-Anna Gatsiou, Irinaios Pilatis, Sophia Sarpaki, George Loudos, Spiros Kostopoulos, Dimitrios Glotsos

Affiliations

  1. Biomedical Engineering Department, University of West Attica, 12210 Athens, Greece.
  2. BIOEMTECH, Lefkippos Attica Technology Park, N.C.S.R. Democritos, 15343 Athens, Greece.

PMID: 34940729 PMCID: PMC8704599 DOI: 10.3390/jimaging7120262

Abstract

In the current work, a pix2pix conditional generative adversarial network has been evaluated as a potential solution for generating adequately accurate synthesized morphological X-ray images by translating standard photographic images of mice. Such an approach will benefit 2D functional molecular imaging techniques, such as planar radioisotope and/or fluorescence/bioluminescence imaging, by providing high-resolution information for anatomical mapping, but not for diagnosis, using conventional photographic sensors. Planar functional imaging offers an efficient alternative to biodistribution ex vivo studies and/or 3D high-end molecular imaging systems since it can be effectively used to track new tracers and study the accumulation from zero point in time post-injection. The superimposition of functional information with an artificially produced X-ray image may enhance overall image information in such systems without added complexity and cost. The network has been trained in 700 input (photography)/ground truth (X-ray) paired mouse images and evaluated using a test dataset composed of 80 photographic images and 80 ground truth X-ray images. Performance metrics such as peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM) and Fréchet inception distance (FID) were used to quantitatively evaluate the proposed approach in the acquired dataset.

Keywords: PET; SPECT; X-ray; cGAN; deep learning; image-to-image translation; molecular preclinical imaging; pix2pix

References

  1. Mol Imaging Biol. 2018 Oct;20(5):716-731 - PubMed
  2. Mol Imaging Biol. 2011 Oct;13(5):949-61 - PubMed
  3. IEEE Trans Radiat Plasma Med Sci. 2021 Mar;5(2):185-192 - PubMed
  4. Clin Radiol. 2010 Jul;65(7):500-16 - PubMed
  5. Nat Rev Drug Discov. 2008 Jul;7(7):591-607 - PubMed
  6. J Control Release. 2020 Jan 10;317:34-42 - PubMed
  7. Nat Rev Drug Discov. 2010 Mar;9(3):203-14 - PubMed
  8. J Imaging. 2021 Aug 04;7(8): - PubMed
  9. Med Phys. 2019 Aug;46(8):3555-3564 - PubMed
  10. Hell J Nucl Med. 2017 May-Aug;20(2):146-153 - PubMed
  11. IEEE Trans Image Process. 2004 Apr;13(4):600-12 - PubMed
  12. Pharmaceutics. 2011 Apr 26;3(2):229-74 - PubMed
  13. Radiol Phys Technol. 2019 Sep;12(3):235-248 - PubMed
  14. Mol Imaging Biol. 2017 Jun;19(3):398-407 - PubMed

Publication Types

Grant support