EUVIMEDEUVIMED entdecken
Crossref

Generative deep learning synthesizes high signal-to-noise ratio sensitivity maps for PET from low count direct normalization data

Mojtaba Jafaritadi iD, Andrew Groll iD, Myungheon Chin iD, Garry Chinn, Jonathan Fisher, Derek Innes, Craig S Levin

Physics in Medicine & Biology · 2026

DOI: 10.1088/1361-6560/ae3ec6 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Physics in Medicine & Biology, 2026. Verfasst von Mojtaba Jafaritadi und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiology, Stanford University, Stanford, CA, United States of America. Hauptthemen laut MEDLINE: Signal-To-Noise Ratio, Positron-Emission Tomography, Image Processing, Computer-Assisted, Deep Learning. Die Arbeit nennt 38 Literaturangaben. Gefördert von: National Institutes of Health, Suomen Kulttuurirahasto.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objective. An accurate and precise normalization procedure is essential to correct for variations in detector efficiency in reconstructed positron emission tomography (PET) images. Direct normalization is a conventional approach that requires a large number of counts per line of response from a known normalization source, which is time-consuming due to the need to acquire very high statistics with a reasonable source strength that does not saturate the system. Approach. To address the challenge of acquiring high signal-to-noise ratio (SNR) PET sensitivity maps efficiently, particularly with the often relatively low-count direct normalization data, this work develops a novel PET data processing and image reconstruction pipeline. This framework integrates sensitivity map features with generative modeling to synthesize high-quality maps, significantly reducing acquisition time while ensuring accurate and efficient normalization. Key contributions comprise a conditional attention-guided generative adversarial network that preserves the geometric and detector-specific characteristics of sensitivity maps, a robust assessment framework to verify synthesized map plausibility, and a comprehensive evaluation of the model’s performance across a range of acquisition and scanner conditions. Main Results. Quantitative evaluations were performed by testing the model on totally unseen normalization data, acquired to reconstruct images of a Hoffman brain phantom, a contrast phantom, and a uniform cylinder phantom. This evaluation used high-count, low-count (1%–15% of high count scan), and synthetic high-count sensitivity maps. The Hoffman brain image volume normalized using a synthetic sensitivity map with 15% count statistics as input produced results that closely matched that using the high count normalization data, with peak SNR (PSNR), structural similarity index measure (SSIM), and normalized root mean square error (NRMSE) values (mean ± standard error) of 30.68 ± 0.31, 0.95 ± 0.00, and 0.35 ± 0.00, respectively. In comparison, the unprocessed sensitivity map with 15% count statistics yielded substantially worse PSNR, SSIM, and NRMSE values of 15.93 ± 0.43, 0.54 ± 0.01, and 1.84 ± 0.03, respectively. Significance. This novel, fast, and effective approach enables high SNR direct normalization of PET image volumes through deep learning using synthetic correction factors obtained from a short normalization scan.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Signal-To-Noise RatioPositron-Emission TomographyImage Processing, Computer-AssistedDeep LearningGenerative Artificial IntelligenceGenerative Adversarial Networks

Stichwörter der Autor:innen

direct normalizationgenerative adversarial networkimage reconstructionpositron emission tomography

Autor:innen und Institutionen

  • Department of Radiology, Stanford University, Stanford, CA, United States of America.
  • Department of Electrical Engineering, Stanford University, Stanford, CA, United States of America.
  • Department of Physics, Stanford University, Stanford, CA, United States of America.
  • Department of Bioengineering, Stanford University, Stanford, CA, United States of America.

ORCID-Profile: Mojtaba Jafaritadi, Andrew Groll, Myungheon Chin

Förderung

  • National Institutes of Health
  • Suomen Kulttuurirahasto

Zitieren

APA

Jafaritadi, M., Groll, A., Chin, M., Chinn, G., Fisher, J., Innes, D., & Levin, C. S. (2026). Generative deep learning synthesizes high signal-to-noise ratio sensitivity maps for PET from low count direct normalization data. Physics in Medicine & Biology. https://doi.org/10.1088/1361-6560/ae3ec6

Vancouver

Jafaritadi M, Groll A, Chin M, Chinn G, Fisher J, Innes D, et al. Generative deep learning synthesizes high signal-to-noise ratio sensitivity maps for PET from low count direct normalization data. Physics in Medicine & Biology. 2026. doi:10.1088/1361-6560/ae3ec6

BibTeX
@article{Jafaritadi2026,
  author = {Jafaritadi, Mojtaba and Groll, Andrew and Chin, Myungheon and Chinn, Garry and Fisher, Jonathan and Innes, Derek and Levin, Craig S},
  title = {Generative deep learning synthesizes high signal-to-noise ratio sensitivity maps for PET from low count direct normalization data},
  journal = {Physics in Medicine & Biology},
  year = {2026},
  doi = {10.1088/1361-6560/ae3ec6}
}
Kostenlos starten →