EUVIMEDEUVIMED entdecken
Crossref

Two-stage deep learning framework for the restoration of incomplete-ring PET images

Yeqi Fang, Rong Zhou

Medical & Biological Engineering & Computing · 2026

DOI: 10.1007/s11517-026-03641-w ↗

Kurz gefasst

Fachartikel in der Zeitschrift Medical & Biological Engineering & Computing, 2026. Verfasst von Yeqi Fang und 1 weiteren Autor:innen. Beteiligte Einrichtung: College of Physics, Sichuan University, Chengdu, 610065, China. fangyeqi@stu.scu.edu.cn. Hauptthemen laut MEDLINE: Positron-Emission Tomography, Deep Learning, Image Processing, Computer-Assisted. Die Arbeit nennt 54 Literaturangaben. Gefördert von: Innovative Research Group Project of the National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Positron Emission Tomography (PET) is an important molecular imaging tool widely used in medicine. Traditional PET systems rely on complete detector rings for full angular coverage and reliable data collection. However, incomplete-ring PET scanners have emerged due to hardware failures, cost constraints, or specific clinical needs. Standard reconstruction algorithms often suffer from performance degradation with these systems because of reduced data completeness and geometric inconsistencies. We present a two-stage deep-learning framework that, without incorporating any time-of-flight (TOF) information, restores high-quality images from data with about 50% missing coincidences–double the loss levels previously addressed by CNN-based methods. The pipeline operates in two stages: a projection-domain Attention U-Net first predicts the missing sections of the sinogram by leveraging spatial context from neighbouring slices, after which the completed data are reconstructed with OSEM algorithm and passed to a cascaded U-Net & warm-start diffusion model for image refinement. This module starts the reverse diffusion process from the U-Net coarse prediction rather than pure Gaussian noise. Using 613 simulated brain volumes from real scans (196 healthy brain samples, 217 Alzheimer’s disease samples, and 200 Mild Cognitive Impairment samples), the results show that our model successfully preserves most anatomical structures and tracer distribution features with PSNR of 38.18 to 38.59 dB and SSIM of 0.9904 to 0.9925. Our two-stage deep-learning framework effectively restores high-quality PET images from over 50 % incomplete-ring data, achieving near-complete anatomical fidelity and robust performance without requiring TOF information. Graphical abstract

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Positron-Emission TomographyDeep LearningImage Processing, Computer-AssistedHumansAlgorithmsBrainConvolutional Neural Networks

Stichwörter der Autor:innen

Attention U-NetDiffusion-based refinementIncomplete-ringSinogram restoration

Autor:innen und Institutionen

  • College of Physics, Sichuan University, Chengdu, 610065, China. fangyeqi@stu.scu.edu.cn.
  • College of Physics, Sichuan University, Chengdu, 610065, China.

Förderung

  • Innovative Research Group Project of the National Natural Science Foundation of China

Zitieren

APA

Fang, Y., & Zhou, R. (2026). Two-stage deep learning framework for the restoration of incomplete-ring PET images. Medical & Biological Engineering & Computing. https://doi.org/10.1007/s11517-026-03641-w

Vancouver

Fang Y, Zhou R. Two-stage deep learning framework for the restoration of incomplete-ring PET images. Medical & Biological Engineering & Computing. 2026. doi:10.1007/s11517-026-03641-w

BibTeX
@article{Fang2026,
  author = {Fang, Yeqi and Zhou, Rong},
  title = {Two-stage deep learning framework for the restoration of incomplete-ring PET images},
  journal = {Medical & Biological Engineering & Computing},
  year = {2026},
  doi = {10.1007/s11517-026-03641-w}
}
Kostenlos starten →