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 Jul 15;25(1):268.
doi: 10.1186/s12911-025-03062-z.

Deep learning-based delineation of whole-body organs at risk empowering adaptive radiotherapy

Zi-Hang Chen #  1 , Song-Feng Li #  2 , Ling-Xin Xu #  1 , Meng-Qiu Tian #  2 , Feng Li #  3 , Yu-Xian Yang  1 , Chen-Fei Wu  1 , Guan-Qun Zhou  1 , Li Lin  4 , Yao Lu  5   6 , Ying Sun  7
Affiliations

Deep learning-based delineation of whole-body organs at risk empowering adaptive radiotherapy

Zi-Hang Chen et al. BMC Med Inform Decis Mak. .

Abstract

Background: Accurate delineation of organs at risk (OARs) is crucial for precision radiotherapy. Most previous autosegmentation models were only constructed for single anatomical region without evaluation of dosimetric impact. We aimed to validate the clinical practicability of deep-learning (DL) models for autosegmentation of whole-body OARs with respect to delineation accuracy, clinical acceptance and dosimetric impact.

Methods: OARs in various anatomical regions, including the head and neck, thorax, abdomen, and pelvis, were automatedly delineated by DL models (DLD) and compared to manual delineations (MD) by an experienced radiation oncologist (RO). The geometric performance was evaluated using the Dice similarity coefficient (DSC) and average surface distance (ASD). RO A corrected DLD to create delineations approved in clinical practice (CPD). RO B graded the accuracy of DLD to assess clinical acceptance. The dosimetric impact was determined by assessing the difference in dosimetric parameters for each OAR in the DLD-based radiotherapy plan (Plan_DLD) and the CPD-based radiotherapy plan (Plan_CPD).

Results: The automatic delineation model has a high OAR delineation accuracy, and the median DSCs can reach 0.841 (IQR, 0.791-0.867) in the head and neck OAR, 0.903 (IQR, 0.777-0.932) in thoracic OAR, 0.847 (IQR, 0.834-0.931) in abdominal OAR, 0.916 (IQR, 0.906-0.964) in pelvic OAR. The majority of DL-generated OARs were graded as clinically acceptable with no editing or little editing needed. No significant differences in dosimetric parameters were found by comparing Plan_DLD with Plan_CPD.

Conclusions: For OARs of whole bodily regions, DL-based segmentation is fast; DL models perform sufficiently well for clinical practice with respect to delineation accuracy, clinical accepatance and dosimetric impact.

Keywords: Auto-delineation; Deep learning; Whole-body organs at risk.

PubMed Disclaimer

Conflict of interest statement

Declarations. Ethical approval: The study protocol was reviewed and approved by Sun Yat-sen University Cancer Center Ethics Committee (B2023-173-01). The study adhered to the principles of the Declaration of Helsinki. Consent to participate: Due to the retrospective nature of the study and only deidentified data included, the Sun Yat-sen University Cancer Center Ethics Committee waived the need for written informed consent. Consent for publication: Not applicable. Competing interests: The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Study flow diagram. The construction of DL models of the thorax, abdomen, and pelvis was similar to that of the DL model of the head and neck. NPC, nasopharyngeal carcinoma; DL, deep learning; DLD, delineations generated by deep learning; CPD, delineations used in clinical practice; RO, radiation oncologist; Plan_DLD, treatment plan based on the delineations generated by deep learning; Plan_CPD, treatment plan based on the delineations used in clinical practice
Fig. 2
Fig. 2
Heatmap for the DSC between DL-generated delineations and manual delineations for each OAR of the whole bodily region. DL, deep learning
Fig. 3
Fig. 3
Degree of volumetric revision to which DLD should be revised to meet clinical criteria. DLD, delineations generated by deep learning
Fig. 4
Fig. 4
Dosimetric evaluation results of Plan_DLD and Plan_CPD. Plan_DLD, treatment plan based on the delineations generated by deep learning; Plan_CPD, treatment plan based on the delineations used in clinical practice. The evaluated metrics for each structure are listed in Supplementary Table 3
Fig. 5
Fig. 5
The dose-volume histogram (DVH) of Plan_CPD and Plan_DLD for representative (A) nasopharyngeal carcinoma case; (B) lung cancer case; (C) liver cancer case; (D) rectal cancer case
Fig. 6
Fig. 6
The correlation between the degree of difference in mean dose and mean DSC for each OAR of the whole body region
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Lim JY, Leech M. Use of auto-segmentation in the delineation of target volumes and organs at risk in head and neck. Acta Oncol. 2016;55(7):799–806. - PubMed
    1. Verhaart RF, et al. CT-based patient modeling for head and neck hyperthermia treatment planning: manual versus automatic normal-tissue-segmentation. Radiother Oncol. 2014;111(1):158–63. - PubMed
    1. Brouwer CL, et al. 3D Variation in delineation of head and neck organs at risk. Radiat Oncol. 2012;7:32. - PMC - PubMed
    1. Vinod SK, et al. Uncertainties in volume delineation in radiation oncology: a systematic review and recommendations for future studies. Radiother Oncol. 2016;121(2):169–79. - PubMed
    1. Ashida R, et al. Impact of interfractional anatomical variation and setup correction methods on interfractional dose variation in IMPT and VMAT plans for pancreatic cancer patients: a planning study. J Appl Clin Med Phys. 2020;21(7):49–59. - PMC - PubMed

MeSH terms

LinkOut - more resources