Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 41 min ago
Peer-reviewedCC BY 4.0Journal of Applied Clinical Medical Physics

Test‐time adaptation‐driven patient‐specific auto‐segmentation for CBCT‐guided online adaptive radiotherapy in cervical cancer

  • Zhiqun Wang1
  • Wei Zhang2
  • Xiangyin Meng1
  • Yongguang Liang1
  • Tingting Dong1
  • Jie Zhang1
  • Nan Liu1
  • Lang Yu1
  • Bo Yang1
  • Qichao Zhou2
  • Jie Qiu1
  1. 1Department of Radiation Oncology Chinese Academy of Medical Sciences, Peking Union Medical College Beijing China
  2. 2Manteia Technologies Co., Ltd. Xiamen Fujian China
Read full textDOICite

Abstract

Background: Online adaptive radiotherapy using cone beam computed tomography (CBCT) of the pelvis can account for anatomical variations during cervical cancer treatment. However, the limitations of CBCT and existing auto‐segmentation methods, such as deformable image registration (DIR), hinder contouring and dosimetric accuracy.

Purpose: We propose a test‐time adaptation (TTA)‐driven patient‐specific segmentation framework to enhance the accuracy and workflow efficiency of CBCT‐guided online adaptive radiotherapy in cervical cancer.

Methods: Our retrospective analysis included 20 patients with cervical/endometrial cancer (530 CBCT scans) treated on the Varian Ethos platform. A patient‐specific UNet variant pre‐trained on the CTPelvic1K dataset using Bootstrap Your Own Latent (BYOL) contrastive learning was fine‐tuned on individual reference CT scans. During daily adaptation, the model incorporated TTA with segmentation (cross entropy + Dice loss) and consistency objectives, leveraging prior CBCT/CT images to constrain anatomical variations. Performance was evaluated using the Dice similarity coefficient (DSC), 95th percentile Hausdorff Distance (HD95), and contouring time reduction compared to DIR‐based methods.

Results: The proposed method significantly outperformed DIR, achieving higher DSC (e.g., postoperative clinical target volume (CTV) of the vaginal cuff and upper vaginal region: 0.89 vs. 0.73; definitive CTV of the pelvic lymphatic drainage area: 0.83 vs. 0.71) and lower HD95 (e.g., rectum: 24.9 vs. 30.7 mm) values. TTA further improved the accuracy, particularly for the bowel (DSC: 0.81 vs. 0.76) and target volumes. Clinically, the proposed method reduced the contouring time by 250 s (postoperative) and 230 s (definitive) per fraction (p < 0.01), with reduced time variability. Structures with clear boundaries required one or two fractions for training, whereas motion‐prone organs required four or five fractions.

Conclusions: This study presents the first retrospective demonstration of TTA‐driven patient‐specific segmentation for pelvic CBCT‐guided online adaptive radiotherapy. The proposed framework enhances segmentation accuracy, workflow efficiency, and temporal consistency by addressing the inherent challenges and data scarcity of CBCT through adaptive learning.

Related

  1. Artificial intelligence in image-guided cervical interventional radiotherapy: a systematic review
    Valentina Lancellotta, Angeles Rovirosa, Maria Concetta La Milia +12
    Frontiers in OncologyPeer-reviewedCC BY 4.0Full textApplied physics & instrumentationCrossrefEurope PMCEurope PMCPMC
  2. Cardiac segmentation for ventricular tachycardia radioablation and radiotherapy patients with cardiac implants
    Nicholas Summerfield, Dustin Jacqmin, Chase Ruff +8
    arXivPreprintCC BY 4.0Full textBiological & medical physicsarXiv
  3. A novel automatic real-time motion tracking method in MRI-guided radiotherapy using enhanced Tracking-Learning-Detection framework with automatic segmentation.
    Shengqi Chen, Zilin Wang, Jianrong Dai +6
    Radiation oncology (London, England)Peer-reviewedCC BY-NC-NDFull textApplied physics & instrumentationEurope PMC
  4. Sam-Correction: Fully Adaptive Label Noise Reduction for Medical Image Segmentation
    Takuro Shimaya, Masahiro Saiko
    arXivPreprintarXiv licenceFull textBiological & medical physicsarXiv
  5. Cone-beam CT-based synthetic CT generation for adaptive radiotherapy: methodological paradigms, clinical reliability assessment, and translation
    Yunxian Zhang, Wanting Zhang, Jinbai Huang +4
    Frontiers in OncologyPeer-reviewedCC BY 4.0Full textApplied physics & instrumentationCrossrefEurope PMCPMC

Cite

BibTeXRIS
BibTeX
@article{wang2026testtime,
  title = {{Test‐time adaptation‐driven patient‐specific auto‐segmentation for CBCT‐guided online adaptive radiotherapy in cervical cancer}},
  author = {Zhiqun Wang and Wei Zhang and Xiangyin Meng and Yongguang Liang and Tingting Dong and Jie Zhang and Nan Liu and Lang Yu and Bo Yang and Qichao Zhou and Jie Qiu},
  journal = {Journal of Applied Clinical Medical Physics},
  year = {2026},
  month = may,
  doi = {10.1002/acm2.70609},
  url = {https://doi.org/10.1002/acm2.70609},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - Test‐time adaptation‐driven patient‐specific auto‐segmentation for CBCT‐guided online adaptive radiotherapy in cervical cancer
AU  - Zhiqun Wang
AU  - Wei Zhang
AU  - Xiangyin Meng
AU  - Yongguang Liang
AU  - Tingting Dong
AU  - Jie Zhang
AU  - Nan Liu
AU  - Lang Yu
AU  - Bo Yang
AU  - Qichao Zhou
AU  - Jie Qiu
PY  - 2026
DA  - 2026/05/14
JO  - Journal of Applied Clinical Medical Physics
DO  - 10.1002/acm2.70609
UR  - https://doi.org/10.1002/acm2.70609
AB  - Background: Online adaptive radiotherapy using cone beam computed tomography (CBCT) of the pelvis can account for anatomical variations during cervical cancer treatment. However, the limitations of CBCT and existing auto‐segmentation methods, such as deformable image registration (DIR), hinder contouring and dosimetric accuracy.  Purpose: We propose a test‐time adaptation (TTA)‐driven patient‐specific segmentation framework to enhance the accuracy and workflow efficiency of CBCT‐guided online adaptive radiotherapy in cervical cancer.  Methods: Our retrospective analysis included 20 patients with cervical/endometrial cancer (530 CBCT scans) treated on the Varian Ethos platform. A patient‐specific UNet variant pre‐trained on the CTPelvic1K dataset using Bootstrap Your Own Latent (BYOL) contrastive learning was fine‐tuned on individual reference CT scans. During daily adaptation, the model incorporated TTA with segmentation (cross entropy + Dice loss) and consistency objectives, leveraging prior CBCT/CT images to constrain anatomical variations. Performance was evaluated using the Dice similarity coefficient (DSC), 95th percentile Hausdorff Distance (HD95), and contouring time reduction compared to DIR‐based methods.  Results: The proposed method significantly outperformed DIR, achieving higher DSC (e.g., postoperative clinical target volume (CTV) of the vaginal cuff and upper vaginal region: 0.89 vs. 0.73; definitive CTV of the pelvic lymphatic drainage area: 0.83 vs. 0.71) and lower HD95 (e.g., rectum: 24.9 vs. 30.7 mm) values. TTA further improved the accuracy, particularly for the bowel (DSC: 0.81 vs. 0.76) and target volumes. Clinically, the proposed method reduced the contouring time by 250 s (postoperative) and 230 s (definitive) per fraction (p < 0.01), with reduced time variability. Structures with clear boundaries required one or two fractions for training, whereas motion‐prone organs required four or five fractions.  Conclusions: This study presents the first retrospective demonstration of TTA‐driven patient‐specific segmentation for pelvic CBCT‐guided online adaptive radiotherapy. The proposed framework enhances segmentation accuracy, workflow efficiency, and temporal consistency by addressing the inherent challenges and data scarcity of CBCT through adaptive learning.
SN  - 1526-9914
ER  -