Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Dec;18(4):1192-1198.
doi: 10.1007/s12194-025-00967-2. Epub 2025 Sep 24.

Dose reduction in radiotherapy treatment planning CT via deep learning-based reconstruction: a single‑institution study

Keisuke Yasui  1 , Yuri Kasugai  2 , Maho Morishita  3 , Yasunori Saito  2 , Hidetoshi Shimizu  4 , Haruka Uezono  5 , Naoki Hayashi  4
Affiliations

Dose reduction in radiotherapy treatment planning CT via deep learning-based reconstruction: a single‑institution study

Keisuke Yasui et al. Radiol Phys Technol. 2025 Dec.

Abstract

To quantify radiation dose reduction in radiotherapy treatment-planning CT (RTCT) using a deep learning-based reconstruction (DLR; AiCE) algorithm compared with adaptive iterative dose reduction (IR; AIDR). To evaluate its potential to inform RTCT-specific diagnostic reference levels (DRLs). In this single-institution retrospective study, 4-part RTCT scans (head, head and neck, lung, and pelvis) were acquired on a large-bore CT. Scans reconstructed with IR (n = 820) and DLR (n = 854) were compared. The 75th-percentile CTDIvol and DLP (CTDIIR, DLPIR vs. CTDIDLR, DLPDLR) were determined per site. Dose reduction rates were calculated as (CTDIDLR - CTDIIR)/CTDIIR × 100% and similarly for DLP. Statistical significance was assessed by the Mann-Whitney U-test. DLR yielded CTDIvol reductions of 30.4-75.4% and DLP reductions of 23.1-73.5% across sites (p < 0.001), with the greatest reductions in head and neck RTCT (CTDIvol: 75.4%; DLP: 73.5%). Variability also narrowed. Compared with published national DRLs, DLR achieved 34.8 mGy and 18.8 mGy lower CTDIvol for head and neck versus UK-DRLs and Japanese multi-institutional data, respectively. DLR substantially lowers RTCT dose indices, providing quantitative data to guide RTCT-specific DRLs and optimize clinical workflows.

Keywords: AiCE; CT dose index; CT reconstruction algorithm; Deep learning-based reconstruction; Treatment planning.

PubMed Disclaimer

Conflict of interest statement

Declarations. Conflict of interest: None. Ethical approval: Ethical approval was obtained from the Fujita Health University IRB (HM24-518), and patient informed consent was waived.

References

    1. Xing L, Thorndyke B, Schreibmann E, Yang Y, Li T-F, Kim G-Y, et al. Overview of image-guided radiation therapy. Med Dosim. 2006;31(2):91–112. - DOI - PubMed
    1. Morin O, Gillis A, Chen J, Aubin M, Bucci MK, Roach M, et al. Megavoltage cone-beam CT: system description and clinical applications. Med Dosim. 2006;31:51–61. - DOI - PubMed
    1. Kidd EA, Siegel BA, Dehdashti F, Rader JS, Mutic S, Mutch DG, et al. Clinical outcomes of definitive intensity-modulated radiation therapy with fluorodeoxyglucose-positron emission tomography simulation in patients with locally advanced cervical cancer. Int J Radiat Oncol Biol Phys. 2010;77:1085–91. - DOI - PubMed
    1. Kirkpatrick JP, Kelsey CR, Palta M, Cabrera AR, Salama JK, Patel P, et al. Stereotactic body radiotherapy: a critical review for nonradiation oncologists: stereotactic body radiotherapy. Cancer. 2014;120:942–54. - DOI - PubMed
    1. ICRP. The 2007 recommendations of the international commission on Radiological Protection. ICRP publication 103. Ann ICRP. 2007;37:1–332.