Cascaded deep learning for synthesis of 18F-FDG PET from 18F-octreotide: model parameter optimization and performance evaluation
- 1PET/CT Center, The First Affiliated Hospital of Xi’an Jiaotong University , ,
Abstract
Objective: To investigate a cascaded conditional generative adversarial network (cGAN)–based method for synthesizing 18F-FDG PET images from 18F-octreotide, optimize model parameter configurations, and evaluate the quality of the synthesized images.
Methods: Imaging data from 71 tumor patients who underwent dual-tracer PET/CT examinations with 18F-FDG and 18F-octreotide between May 2022 and November 2025 were retrospectively collected. A cascaded deep learning architecture (first stage: 3D U-Net; second stage: self-attention 3D V-Net) was employed to evaluate the effects of image preprocessing, resampling size, and patch size on synthesis quality. Three preprocessing strategies (log transformation, normalization, and a combined approach) were compared. Image quality was assessed using five objective metrics (SSIM, PSNR, MSE, MAE, and NMI) and Likert 5-point ratings from three nuclear medicine physicians; liver and lesion-level SUV agreement were further evaluated.
Results: In the held-out test cohort (n = 14), the combined preprocessing method (log transformation + normalization) significantly outperformed the individual methods (SSIM: 0.8454 ± 0.0328 vs. 0.7831 ± 0.0337 vs. 0.7438 ± 0.0362, P < 0.001). On the identical held-out test cohort (n = 14), the Group 2 parameter configuration (resampling size: 256 × 256 × 512; patch size: 128 × 128 × 128) outperformed Group 1, with an SSIM of 0.8854 ± 0.0455 and a PSNR of 28.15 ± 3.35 dB, representing relative improvements of 19.1% and 83.5%, respectively, over Group 1 (paired-samples t-test, both P < 0.001). The mean Likert score was 4.53 ± 0.19. Quantitative analysis of liver SUVmean showed no significant difference between synthesized and real images (P = 0.967, paired t-test). Lesion-level quantitative analysis demonstrated high agreement between synthetic and real 18F-FDG images, with Pearson correlation coefficients of 1.0000, 0.9995, and 0.9991 for SUVmax and SUVmean (n = 14) and SUVpeak (n = 13), respectively (all P < 0.001), no significant differences on paired t-tests (all P > 0.05), and all Bland–Altman differences within the 95% limits of agreement.
Conclusion: The cascaded deep learning model enabled cross-tracer synthesis from 18F-octreotide to 18F-FDG PET images. In this exploratory, single-center pilot cohort, the optimized configuration demonstrated promising quantitative and visual performance, supporting image-synthesis feasibility and potential clinical utility of a single-injection, dual-parameter molecular imaging approach for neuroendocrine tumors. Prospective multicenter validation is warranted before any clinical application.
Related
- Performance evaluation and optimization of powder mixed EDM of Inconel 718 using composite electrodes and the VIKOR method
- Oxalate Nephropathy in a patient with Neuroendocrine tumor receiving Long-acting Octreotide
- Evaluating parameter-efficient fine-tuning of large language models for sentiment analysis with bayesian optimization
- Case Report: Neurosurgery, octreotide, cabergoline, radiotherapy, pasireotide and pegvisomant – a multimodal strategy for adolescent acrogigantism
- Discrete element simulation and dimensionless parameter-based optimization of near-wellbore fracture propagation during perforation fracturing of tight sandstone
BibTeX
@article{chang2026cascaded,
title = {{Cascaded deep learning for synthesis of 18F-FDG PET from 18F-octreotide: model parameter optimization and performance evaluation}},
author = {Ruxi Chang and Liping Yang and Jungang Gao and Xiang Liu and Bo Wang and Yan Li and Xiaoyi Duan},
journal = {Frontiers in nuclear medicine},
year = {2026},
month = oct,
doi = {10.3389/fnume.2026.1920721},
url = {https://europepmc.org/article/PMC/PMC13719449},
note = {CC BY},
}
RIS
TY - JOUR
TI - Cascaded deep learning for synthesis of 18F-FDG PET from 18F-octreotide: model parameter optimization and performance evaluation
AU - Ruxi Chang
AU - Liping Yang
AU - Jungang Gao
AU - Xiang Liu
AU - Bo Wang
AU - Yan Li
AU - Xiaoyi Duan
PY - 2026
DA - 2026/10/16
JO - Frontiers in nuclear medicine
DO - 10.3389/fnume.2026.1920721
UR - https://europepmc.org/article/PMC/PMC13719449
AB - Objective: To investigate a cascaded conditional generative adversarial network (cGAN)–based method for synthesizing 18F-FDG PET images from 18F-octreotide, optimize model parameter configurations, and evaluate the quality of the synthesized images. Methods: Imaging data from 71 tumor patients who underwent dual-tracer PET/CT examinations with 18F-FDG and 18F-octreotide between May 2022 and November 2025 were retrospectively collected. A cascaded deep learning architecture (first stage: 3D U-Net; second stage: self-attention 3D V-Net) was employed to evaluate the effects of image preprocessing, resampling size, and patch size on synthesis quality. Three preprocessing strategies (log transformation, normalization, and a combined approach) were compared. Image quality was assessed using five objective metrics (SSIM, PSNR, MSE, MAE, and NMI) and Likert 5-point ratings from three nuclear medicine physicians; liver and lesion-level SUV agreement were further evaluated. Results: In the held-out test cohort (n = 14), the combined preprocessing method (log transformation + normalization) significantly outperformed the individual methods (SSIM: 0.8454 ± 0.0328 vs. 0.7831 ± 0.0337 vs. 0.7438 ± 0.0362, P < 0.001). On the identical held-out test cohort (n = 14), the Group 2 parameter configuration (resampling size: 256 × 256 × 512; patch size: 128 × 128 × 128) outperformed Group 1, with an SSIM of 0.8854 ± 0.0455 and a PSNR of 28.15 ± 3.35 dB, representing relative improvements of 19.1% and 83.5%, respectively, over Group 1 (paired-samples t-test, both P < 0.001). The mean Likert score was 4.53 ± 0.19. Quantitative analysis of liver SUVmean showed no significant difference between synthesized and real images (P = 0.967, paired t-test). Lesion-level quantitative analysis demonstrated high agreement between synthetic and real 18F-FDG images, with Pearson correlation coefficients of 1.0000, 0.9995, and 0.9991 for SUVmax and SUVmean (n = 14) and SUVpeak (n = 13), respectively (all P < 0.001), no significant differences on paired t-tests (all P > 0.05), and all Bland–Altman differences within the 95% limits of agreement. Conclusion: The cascaded deep learning model enabled cross-tracer synthesis from 18F-octreotide to 18F-FDG PET images. In this exploratory, single-center pilot cohort, the optimized configuration demonstrated promising quantitative and visual performance, supporting image-synthesis feasibility and potential clinical utility of a single-injection, dual-parameter molecular imaging approach for neuroendocrine tumors. Prospective multicenter validation is warranted before any clinical application.
SN - 2673-8880
ER -