18F-FDG dose reduction using deep learning-based PET reconstruction

Abstract Background A deep learning-based image reconstruction (DLR) algorithm that can reduce the statistical noise has been developed for PET/CT imaging. It may reduce the administered dose of 18F-FDG and minimize radiation exposure while maintaining diagnostic quality. This retrospective study ev...

Full description

Saved in:
Bibliographic Details
Main Authors: Ryuji Akita, Komei Takauchi, Mana Ishibashi, Shota Kondo, Shogo Ono, Kazushi Yokomachi, Yusuke Ochi, Masao Kiguchi, Hidenori Mitani, Yuko Nakamura, Kazuo Awai
Format: Article
Language:English
Published: SpringerOpen 2025-07-01
Series:EJNMMI Research
Subjects:
Online Access:https://doi.org/10.1186/s13550-025-01269-9
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849238270376935424
author Ryuji Akita
Komei Takauchi
Mana Ishibashi
Shota Kondo
Shogo Ono
Kazushi Yokomachi
Yusuke Ochi
Masao Kiguchi
Hidenori Mitani
Yuko Nakamura
Kazuo Awai
author_facet Ryuji Akita
Komei Takauchi
Mana Ishibashi
Shota Kondo
Shogo Ono
Kazushi Yokomachi
Yusuke Ochi
Masao Kiguchi
Hidenori Mitani
Yuko Nakamura
Kazuo Awai
author_sort Ryuji Akita
collection DOAJ
description Abstract Background A deep learning-based image reconstruction (DLR) algorithm that can reduce the statistical noise has been developed for PET/CT imaging. It may reduce the administered dose of 18F-FDG and minimize radiation exposure while maintaining diagnostic quality. This retrospective study evaluated whether the injected 18F-FDG dose could be reduced by applying DLR to PET images. To this aim, we compared the quantitative image quality metrics and the false-positive rate between DLR with a reduced 18F-FDG dose and Ordered Subsets Expectation Maximization (OSEM) with a standard dose. Results This study included 90 oncology patients who underwent 18F-FDG PET/CT. They were divided into 3 groups (30 patients each): group A (18F-FDG dose per body weight [BW]: 2.00—2.99 MBq/kg; PET image reconstruction: DLR), group B (3.00–3.99 MBq/kg; DLR), and group C (standard dose group; 4.00—4.99 MBq/kg; OSEM). The evaluation was performed using the signal-to-noise ratio (SNR), target-to-background ratio (TBR), and false-positive rate. DLR yielded significantly higher SNRs in groups A and B than group C (p < 0.001). There was no significant difference in the TBR between groups A and C, and between groups B and C (p = 0.983 and 0.605, respectively). In group B, more than 80% of patients weighing less than 75 kg had at most one false positive result. In contrast, in group B patients weighing 75 kg or more, as well as in group A, less than 80% of patients had at most one false-positives. Conclusions Our findings suggest that the injected 18F-FDG dose can be reduced to 3.0 MBq/kg in patients weighing less than 75 kg by applying DLR. Compared to the recommended dose in the European Association of Nuclear Medicine (EANM) guidelines for 90 s per bed position (4.7 MBq/kg), this represents a dose reduction of 36%. Further optimization of DLR algorithms is required to maintain comparable diagnostic accuracy in patients weighing 75 kg or more.
format Article
id doaj-art-ff3ffd66642e4a95bc3714142c9b45e7
institution Kabale University
issn 2191-219X
language English
publishDate 2025-07-01
publisher SpringerOpen
record_format Article
series EJNMMI Research
spelling doaj-art-ff3ffd66642e4a95bc3714142c9b45e72025-08-20T04:01:41ZengSpringerOpenEJNMMI Research2191-219X2025-07-011511910.1186/s13550-025-01269-918F-FDG dose reduction using deep learning-based PET reconstructionRyuji Akita0Komei Takauchi1Mana Ishibashi2Shota Kondo3Shogo Ono4Kazushi Yokomachi5Yusuke Ochi6Masao Kiguchi7Hidenori Mitani8Yuko Nakamura9Kazuo Awai10Department of Diagnostic Radiology, Graduate School of Biomedical and Health Sciences, Hiroshima UniversityDepartment of Radiology, Hiroshima University HospitalCenter for Radiation Disaster Medical Science, Research Institute for Radiation Biology and Medicine, Hiroshima UniversityDepartment of Diagnostic Radiology, Graduate School of Biomedical and Health Sciences, Hiroshima UniversityCT Systems Development Department, Canon Medical Systems CorporationDepartment of Radiology, Hiroshima University HospitalDepartment of Radiology, Hiroshima University HospitalDepartment of Radiology, Hiroshima University HospitalDepartment of Diagnostic Radiology, Graduate School of Biomedical and Health Sciences, Hiroshima UniversityDepartment of Diagnostic Radiology, Graduate School of Biomedical and Health Sciences, Hiroshima UniversityDepartment of Diagnostic Radiology, Graduate School of Biomedical and Health Sciences, Hiroshima UniversityAbstract Background A deep learning-based image reconstruction (DLR) algorithm that can reduce the statistical noise has been developed for PET/CT imaging. It may reduce the administered dose of 18F-FDG and minimize radiation exposure while maintaining diagnostic quality. This retrospective study evaluated whether the injected 18F-FDG dose could be reduced by applying DLR to PET images. To this aim, we compared the quantitative image quality metrics and the false-positive rate between DLR with a reduced 18F-FDG dose and Ordered Subsets Expectation Maximization (OSEM) with a standard dose. Results This study included 90 oncology patients who underwent 18F-FDG PET/CT. They were divided into 3 groups (30 patients each): group A (18F-FDG dose per body weight [BW]: 2.00—2.99 MBq/kg; PET image reconstruction: DLR), group B (3.00–3.99 MBq/kg; DLR), and group C (standard dose group; 4.00—4.99 MBq/kg; OSEM). The evaluation was performed using the signal-to-noise ratio (SNR), target-to-background ratio (TBR), and false-positive rate. DLR yielded significantly higher SNRs in groups A and B than group C (p < 0.001). There was no significant difference in the TBR between groups A and C, and between groups B and C (p = 0.983 and 0.605, respectively). In group B, more than 80% of patients weighing less than 75 kg had at most one false positive result. In contrast, in group B patients weighing 75 kg or more, as well as in group A, less than 80% of patients had at most one false-positives. Conclusions Our findings suggest that the injected 18F-FDG dose can be reduced to 3.0 MBq/kg in patients weighing less than 75 kg by applying DLR. Compared to the recommended dose in the European Association of Nuclear Medicine (EANM) guidelines for 90 s per bed position (4.7 MBq/kg), this represents a dose reduction of 36%. Further optimization of DLR algorithms is required to maintain comparable diagnostic accuracy in patients weighing 75 kg or more.https://doi.org/10.1186/s13550-025-01269-918F-FDGPET/CTDose reductionDeep learning-based reconstructionImage quality
spellingShingle Ryuji Akita
Komei Takauchi
Mana Ishibashi
Shota Kondo
Shogo Ono
Kazushi Yokomachi
Yusuke Ochi
Masao Kiguchi
Hidenori Mitani
Yuko Nakamura
Kazuo Awai
18F-FDG dose reduction using deep learning-based PET reconstruction
EJNMMI Research
18F-FDG
PET/CT
Dose reduction
Deep learning-based reconstruction
Image quality
title 18F-FDG dose reduction using deep learning-based PET reconstruction
title_full 18F-FDG dose reduction using deep learning-based PET reconstruction
title_fullStr 18F-FDG dose reduction using deep learning-based PET reconstruction
title_full_unstemmed 18F-FDG dose reduction using deep learning-based PET reconstruction
title_short 18F-FDG dose reduction using deep learning-based PET reconstruction
title_sort 18f fdg dose reduction using deep learning based pet reconstruction
topic 18F-FDG
PET/CT
Dose reduction
Deep learning-based reconstruction
Image quality
url https://doi.org/10.1186/s13550-025-01269-9
work_keys_str_mv AT ryujiakita 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT komeitakauchi 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT manaishibashi 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT shotakondo 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT shogoono 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT kazushiyokomachi 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT yusukeochi 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT masaokiguchi 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT hidenorimitani 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT yukonakamura 18ffdgdosereductionusingdeeplearningbasedpetreconstruction
AT kazuoawai 18ffdgdosereductionusingdeeplearningbasedpetreconstruction