EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Localized fine-tuning and clinical evaluation of deep-learning based auto-segmentation (DLAS) model for clinical target volume (CTV) and organs-at-risk (OAR) in rectal cancer radiotherapy

Jianhao Geng, Xin Sui, Rongxu Du, Jialin Feng, Ruoxi Wang, Meijiao Wang, Kaining Yao, Qi Chen, Lu Bai, Shaobin Wang, Yongheng Li, Hao Wu und 2 weitere

Radiation Oncology · 2024

DOI: 10.1186/s13014-024-02463-0 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Radiation Oncology, 2024. Verfasst von Jianhao Geng und 13 weiteren Autor:innen. Beteiligte Einrichtung: Department of Radiation Oncology, Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Peking University Canc…. Hauptthemen laut MEDLINE: Rectal Neoplasms, Deep Learning, Organs at Risk, Radiotherapy Planning, Computer-Assisted. Die Arbeit nennt 25 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background and purpose Various deep learning auto-segmentation (DLAS) models have been proposed, some of which have been commercialized. However, the issue of performance degradation is notable when pretrained models are deployed in the clinic. This study aims to enhance precision of a popular commercial DLAS product in rectal cancer radiotherapy by localized fine-tuning, addressing challenges in practicality and generalizability in real-world clinical settings. Materials and methods A total of 120 Stage II/III mid-low rectal cancer patients were retrospectively enrolled and divided into three datasets: training ( n = 60), external validation (ExVal, n = 30), and generalizability evaluation (GenEva, n = 30) datasets respectively. The patients in the training and ExVal dataset were acquired on the same CT simulator, while those in GenEva were on a different CT simulator. The commercial DLAS software was first localized fine-tuned (LFT) for clinical target volume (CTV) and organs-at-risk (OAR) using the training data, and then validated on ExVal and GenEva respectively. Performance evaluation involved comparing the LFT model with the vendor-provided pretrained model (VPM) against ground truth contours, using metrics like Dice similarity coefficient (DSC), 95th Hausdorff distance (95HD), sensitivity and specificity.

Results

LFT significantly improved CTV delineation accuracy ( p < 0.05) with LFT outperforming VPM in target volume, DSC, 95HD and specificity. Both models exhibited adequate accuracy for bladder and femoral heads, and LFT demonstrated significant enhancement in segmenting the more complex small intestine. We did not identify performance degradation when LFT and VPM models were applied in the GenEva dataset.

Conclusions

The necessity and potential benefits of LFT DLAS towards institution-specific model adaption is underscored. The commercial DLAS software exhibits superior accuracy once localized fine-tuned, and is highly robust to imaging equipment changes.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Rectal NeoplasmsDeep LearningOrgans at RiskRadiotherapy Planning, Computer-AssistedHumansRetrospective StudiesFemaleMaleMiddle AgedAgedRadiotherapy DosageTomography, X-Ray ComputedAdultRadiotherapy, Intensity-Modulated

Stichwörter der Autor:innen

Auto-contouringClinical target volumeDeep learningOrgan at riskRectal cancer

Autor:innen und Institutionen

  • Department of Radiation Oncology, Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Peking University Cancer Hospital & Institute, Beijing, 100142, China.
  • Research and Development Department, MedMind Technology Co., Ltd, Beijing, 100083, China.
  • Institute of Medical Technology, Peking University Health Science Center, Beijing, 100191, China.
  • Beijing Key Lab of Nanophotonics and Ultrafine Optoelectronic Systems, School of Optics and Photonics, Beijing Institute of Technology, Beijing, 100081, China. hu.xiangmin@bit.edu.cn.
  • Department of Radiation Oncology, Key Laboratory of Carcinogenesis and Translational Research (Ministry of Education/Beijing), Peking University Cancer Hospital & Institute, Beijing, 100142, China. yidu_rt@163.com.
  • Institute of Medical Technology, Peking University Health Science Center, Beijing, 100191, China. yidu_rt@163.com.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Geng, J., Sui, X., Du, R., Feng, J., Wang, R., Wang, M., Yao, K., Chen, Q., Bai, L., Wang, S., Li, Y., Wu, H., Hu, X., & Du, Y. (2024). Localized fine-tuning and clinical evaluation of deep-learning based auto-segmentation (DLAS) model for clinical target volume (CTV) and organs-at-risk (OAR) in rectal cancer radiotherapy. Radiation Oncology. https://doi.org/10.1186/s13014-024-02463-0

Vancouver

Geng J, Sui X, Du R, Feng J, Wang R, Wang M, et al. Localized fine-tuning and clinical evaluation of deep-learning based auto-segmentation (DLAS) model for clinical target volume (CTV) and organs-at-risk (OAR) in rectal cancer radiotherapy. Radiation Oncology. 2024. doi:10.1186/s13014-024-02463-0

BibTeX
@article{Geng2024,
  author = {Geng, Jianhao and Sui, Xin and Du, Rongxu and Feng, Jialin and Wang, Ruoxi and Wang, Meijiao and Yao, Kaining and Chen, Qi and Bai, Lu and Wang, Shaobin and Li, Yongheng and Wu, Hao and Hu, Xiangmin and Du, Yi},
  title = {Localized fine-tuning and clinical evaluation of deep-learning based auto-segmentation (DLAS) model for clinical target volume (CTV) and organs-at-risk (OAR) in rectal cancer radiotherapy},
  journal = {Radiation Oncology},
  year = {2024},
  doi = {10.1186/s13014-024-02463-0}
}
Kostenlos starten →