Scholarlink

Background <p>Accurate PET quantification relies on attenuation correction (AC), commonly performed using a linear attenuation map derived from a CT acquisition. However, CT can introduce misregistration artifacts and adds radiation dose. Synthetic CT (sCT) from non-attenuation corrected (NAC) PET offers a CT-less alternative, but training robust models requires multi-site data that may be difficult to share under privacy regulations. We aim to enable PET-based sCT training across sites without exposing data. </p> Methods <p> We built a federated learning (FL) framework and trained two sCT generators–a paired conditional GAN and a CycleGAN. Models were pretrained on a single-site cohort (Site&#xa0;1, <i>n</i> = 425) and fine-tuned via FL using additional data from Site&#xa0;1 (<i>n</i> = 25) and a second site with different scanners and reconstruction parameters (Site&#xa0;2, <i>n</i> = 25). Performance was assessed on an internal hold-out set (Site&#xa0;1, <i>n</i> = 91) and two external cohorts (Sites&#xa0;3,4; <i>n</i> = 11, 10) using region-wise relative mean error (rME) of SUV in AC PET. </p> Results <p>Both models produced anatomically plausible sCT and AC PET with low errors when test data matched training distributions. FL fine-tuning improved robustness under distribution shift at Site&#xa0;3, reducing errors across most regions, while maintaining comparable performance at Site&#xa0;4 where protocols resembled the pretraining site. </p> Conclusion <p> Multi-site FL is a feasible path to increase the generalizability of PET-based sCT while preserving data privacy. The proposed framework offers a practical template for training and deploying CT-less AC models across heterogeneous clinical environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multi-site distributed training with data protections for PET-based synthetic CT

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年4月15日

查看全文
  • Kasper Jørgensen,
  • Lauren Partin,
  • Raghavan Ashok,
  • Vijay Shah,
  • Anders Bertil Rodell,
  • Martin Bazik,
  • Bruce Spottiswoode,
  • Flemming Littrup Andersen

摘要

Background

Accurate PET quantification relies on attenuation correction (AC), commonly performed using a linear attenuation map derived from a CT acquisition. However, CT can introduce misregistration artifacts and adds radiation dose. Synthetic CT (sCT) from non-attenuation corrected (NAC) PET offers a CT-less alternative, but training robust models requires multi-site data that may be difficult to share under privacy regulations. We aim to enable PET-based sCT training across sites without exposing data.

Methods

We built a federated learning (FL) framework and trained two sCT generators–a paired conditional GAN and a CycleGAN. Models were pretrained on a single-site cohort (Site 1, n = 425) and fine-tuned via FL using additional data from Site 1 (n = 25) and a second site with different scanners and reconstruction parameters (Site 2, n = 25). Performance was assessed on an internal hold-out set (Site 1, n = 91) and two external cohorts (Sites 3,4; n = 11, 10) using region-wise relative mean error (rME) of SUV in AC PET.

Results

Both models produced anatomically plausible sCT and AC PET with low errors when test data matched training distributions. FL fine-tuning improved robustness under distribution shift at Site 3, reducing errors across most regions, while maintaining comparable performance at Site 4 where protocols resembled the pretraining site.

Conclusion

Multi-site FL is a feasible path to increase the generalizability of PET-based sCT while preserving data privacy. The proposed framework offers a practical template for training and deploying CT-less AC models across heterogeneous clinical environments.