What question did this study set out to answer?

This research investigates how the depth of residual networks affects the performance of PET image restoration under low-count conditions.

April 1, 2026Open Access

Depth-Dependent Performance of Residual Networks for Low-Count PET Image Restoration Using a Dedicated 3D-Printed Striatum Phantom

Key Points

This research investigates how the depth of residual networks affects the performance of PET image restoration under low-count conditions.
Used a 3D-printed striatum phantom for controlled imaging experiments.
Acquired low-count PET images with a clinical PET/CT system.
Compared ResNet-based models with conventional filters (median, Wiener) on restoration quality.
Assessed image quality through various metrics: CNR, COV, UQI, and LPIPS.
ResNet-based methods significantly surpassed conventional filters in contrast recovery and structural integrity.
The ResNet-16 model achieved the highest CNR (9.02) and lowest COV (0.105).
Deeper network configurations offered diminishing returns in quality and structural consistency.

Abstract

Low-count positron emission tomography (PET) is inherently affected by Poisson-dominated noise, which degrades image contrast, structural delineation, and quantitative reliability. This study systematically evaluated residual learning-based deep neural networks to investigate the influence of residual block depth on PET image restoration performance under low-count conditions. We employed a physically controlled striatum phantom, fabricated using 3D printing technology, to ensure reproducible acquisition conditions and controlled physical variability. PET images were acquired using a clinical PET/computed tomography (CT) system with list-mode acquisition. Low-count images reconstructed from short-duration acquisition were paired with high-count reference images reconstructed from extended acquisitions. We compared conventional filtering techniques, including median, Wiener, and modified median Wiener filters, with residual network (ResNet)-based models incorporating 8, 16, and 32 residual blocks. Image quality was quantitatively assessed using contrast-to-noise ratio (CNR), coefficient of variation (COV), line profile analysis, universal quality index (UQI), and perceptual image patch similarity (LPIPS). The results demonstrated that ResNet-based restorations substantially outperformed conventional filtering techniques in contrast recovery, signal stability, and structural preservation. The ResNet-16 model achieved the most balanced performance, yielding the highest CNR (9.02) and lowest COV (0.105), while also demonstrating superior structural and perceptual similarity, as indicated by UQI (0.9224) and LPIPS (0.0174), relative to the high-count reference images. Deeper network configurations exhibited diminishing returns and reduced structural consistencies. These findings indicate that an intermediate residual block depth is optimal for low-count PET image restoration and highlight the importance of architectural optimization in deep learning-based PET image enhancement with phantom-based evaluation frameworks.

Read Full Paperexternally

Bookmark

View Full Paper

Cite This Study

Park et al. (Fri,) studied this question.

synapsesocial.com/papers/69ccb6e416edfba7beb889b2 https://doi.org/https://doi.org/10.3390/bioengineering13040392

Bookmark

View Full Paper