Poster (After Call)

FZJ-2025-01694

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Enhancing Ultra-Low-Dose PET/MRI Using Deep Learning Method for Improved Interpretation

Masood, A. (First author) ; Drzezga, A.FZJ* ; Elmenhorst, E. M. ; Foerges, A. L.FZJ* ; Lange, D. ; Hennecke, E. ; Baur, D. M. ; Kroll, T.FZJ* ; Neumaier, B.FZJ* ; Aeschbach, D. ; Bauer, A.FZJ* ; Landolt, H. P. ; Elmenhorst, D.FZJ* ; Beer, S. (Last author)FZJ*

2024

Abstract: Motivation: We developed a deep learning model to enhance the image quality of ultra-low dose brain PET.Goal(s): Significantly reducing the injected dose not only minimizes radiation risk in subjects but also provides options for scanning protocols, and more follow-up studies.Approach: We proposed a 3D-Residual Attention U-Net model initially trained on whole-body [18F]FDG PET/MR images. We used transfer learning approach to fine-tune our proposed model on [18F]CPFPX PET/MRI inhouse dataset.Results: We achieved improved metrics compared to U-Net model with average PSNR of 28.02 (U-Net: 21.23), SSIM of 0.81 (U-Net: 0.53), CNR of 0.72 (U-Net: 0.61) and NMSE of 0.33 (U-Net: 0.67).Impact: Our model has potential to generate high-quality PET images from low-dose PET/MR, potentially contribute to implementation of kinetic modelling using PET/MR imaging. Our model is capable of enhancing both whole-body and brain datasets, making it valuable asset for diverse applications.

---

Contributing Institute(s):

1. Molekulare Organisation des Gehirns (INM-2)

Research Program(s):

1. 5254 - Neuroscientific Data Analytics and AI (POF4-525) (POF4-525)

Appears in the scientific report 2024

---