EUVIMEDEUVIMED entdecken
Crossref

Machine Learning Models Predicting Radiation Pneumonitis Based on a Multiregional Radiomics and Dosiomics Approach in Patients With Lung Cancer Treated on the NRG/RTOG 0617 Clinical Trial

Taman Upadhaya, Indrin J. Chetty, Bijaya Acharya, Elizabeth M. McKenzie, Hassan Bagher-Ebadian, Katelyn M. Atkins

IEEE Transactions on Radiation and Plasma Medical Sciences · 2026-11 · 10 (8) · 1353-1360

DOI: 10.1109/trpms.2026.3663985 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Radiation and Plasma Medical Sciences, 2026-11. Verfasst von Taman Upadhaya und 5 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 42 Literaturangaben. Gefördert von: Luke Wu-Jei Chang Developmental Funds Award., Cedars-Sinai Radonc, Luke Wu-Jei Chang Developmental Funds.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We explored a multi-regional imaging-omics approach to extract computed tomography (CT)-based radiomics and 3-dimenional (3D) dosiomics features to predict radiation pneumonitis (RP) in patients with locally advanced non-small cell lung cancer (LA-NSCLC). This study analyzed 451 patients from the NRG Oncology/Radiation Therapy Oncology Group (RTOG) 0617 clinical trial data set. A radiomics analysis pipeline was implemented to extract features from CT and 3D dose maps using multi-regional lung contours (clinical target volume; ipsilateral, contralateral and total lungs) to predict grade≥2 RP. Predictive models, generalized linear model with LASSO penalty (LASSO-GLM), Random Forest (RF), XGBoost, and Support Vector Machine (linear and RBF) were evaluated. The model combining radiomics with dosiomics-based features from total lungs achieved the highest area under the curve (AUC, 0.74; 95% CI, 0.65-0.83) on the validation dataset and 0.70 (95% CI, 0.56-0.83) on the test dataset using SVM-RBF algorithm. Combining radiomics, dosiomics, and filter-based features from the contralateral lung both achieved AUC of 0.69 in the test dataset using the SVM-RBF algorithm, similar to dosiomics features extracted from total lungs. These results are suggestive that radiomics, dosiomics, and filter features on radiotherapy planning CT scans and total lung 3D dose distributions can provide complimentary information to enhance model performance and inform signatures for prediction of RP.

Förderung

  • Luke Wu-Jei Chang Developmental Funds Award.
  • Cedars-Sinai Radonc
  • Luke Wu-Jei Chang Developmental Funds
  • OncLive
  • Varian Medical Systems

Zitieren

APA

Upadhaya, T., Chetty, I. J., Acharya, B., McKenzie, E. M., Bagher-Ebadian, H., & Atkins, K. M. (2026). Machine Learning Models Predicting Radiation Pneumonitis Based on a Multiregional Radiomics and Dosiomics Approach in Patients With Lung Cancer Treated on the NRG/RTOG 0617 Clinical Trial. IEEE Transactions on Radiation and Plasma Medical Sciences, 10(8), 1353-1360. https://doi.org/10.1109/trpms.2026.3663985

Vancouver

Upadhaya T, Chetty IJ, Acharya B, McKenzie EM, Bagher-Ebadian H, Atkins KM. Machine Learning Models Predicting Radiation Pneumonitis Based on a Multiregional Radiomics and Dosiomics Approach in Patients With Lung Cancer Treated on the NRG/RTOG 0617 Clinical Trial. IEEE Transactions on Radiation and Plasma Medical Sciences. 2026;10(8):1353-1360. doi:10.1109/trpms.2026.3663985

BibTeX
@article{Upadhaya2026,
  author = {Upadhaya, Taman and Chetty, Indrin J. and Acharya, Bijaya and McKenzie, Elizabeth M. and Bagher-Ebadian, Hassan and Atkins, Katelyn M.},
  title = {Machine Learning Models Predicting Radiation Pneumonitis Based on a Multiregional Radiomics and Dosiomics Approach in Patients With Lung Cancer Treated on the NRG/RTOG 0617 Clinical Trial},
  journal = {IEEE Transactions on Radiation and Plasma Medical Sciences},
  year = {2026},
  volume = {10},
  number = {8},
  pages = {1353--1360},
  doi = {10.1109/trpms.2026.3663985}
}
Kostenlos starten →