Skip to main content

WE‐D‐M100J‐04: Prediction of Lung Radiation‐Induced Pneumonitis Using the Support Vector Machine Algorithm

Publication ,  Journal Article
Chen, S; Zhou, S; Zhang, J; Marks, L; Das, S
Published in: Medical Physics
January 1, 2007

Purpose: To build and test a Support Vector Machine (SVM) model to predict the occurrence of lung radiation‐induced Grade 2+ pneumonitis. SVM is a sophisticated statistical technique that is capable of using complex hypersurfaces to separate the cases with and without pneumonitis. Method and Materials: Two SVM models were built using data from 235 patients with lung cancer treated using radiotherapy (34 diagnosed with pneumonitis). One model (SVMall) selected input features from all dose‐volume and non‐dose factors. For comparison, the other model (SVMdose) selected input features only from lung dose‐volume factors. The models were built with in‐house developed software that employed a unique strategy to sequentially add/remove/substitute features. The SVM models were tested using ten‐fold cross‐validation, wherein 1/10th of the data were tested, in turn, using the model built with the remaining 9/10th of the data. Results: The input features selected to build SVMall were the lung generalized equivalent uniform dose (EUD) with exponents a=1.2, 1.3, 1.4, chemotherapy prior to radiotherapy (yes/no), tumor location (central/peripheral), gender, and histology (adenocarcinoma/other; small cell/other). The input features for SVMdose were EUD a = 1.1, 1.3, 1.4, lung volume receiving > 48 Gy (V48), and V50. Both models selected EUD a ≈ 1 (EUD a=1 is the mean lung dose, which frequently appears as a strong predictor of radiation pneumonitis in literature). The area under the cross‐validated SVMall Receiver Operating Characteristics curve was 0.76 (sensitivity/specificity = 74%/75%), compared to the corresponding SVMdose area of 0.71 (sensitivity/specificity = 68%/68%). SVMall was statistically superior (p=0.01), indicating that non‐dose features significantly contribute to separating patients with and without pneumonitis. Conclusions: The SVM model constructed from dose and non‐dose input factors is a valuable prospective tool for predicting the occurrence of radiation‐induced lung pneumonitis. © 2007, American Association of Physicists in Medicine. All rights reserved.

Duke Scholars

Published In

Medical Physics

DOI

ISSN

0094-2405

Publication Date

January 1, 2007

Volume

34

Issue

6

Start / End Page

2602 / 2603

Related Subject Headings

  • Nuclear Medicine & Medical Imaging
  • 1112 Oncology and Carcinogenesis
  • 0903 Biomedical Engineering
  • 0299 Other Physical Sciences
 

Citation

APA
Chicago
ICMJE
MLA
NLM
Chen, S., Zhou, S., Zhang, J., Marks, L., & Das, S. (2007). WE‐D‐M100J‐04: Prediction of Lung Radiation‐Induced Pneumonitis Using the Support Vector Machine Algorithm. Medical Physics, 34(6), 2602–2603. https://doi.org/10.1118/1.2761565
Chen, S., S. Zhou, J. Zhang, L. Marks, and S. Das. “WE‐D‐M100J‐04: Prediction of Lung Radiation‐Induced Pneumonitis Using the Support Vector Machine Algorithm.” Medical Physics 34, no. 6 (January 1, 2007): 2602–3. https://doi.org/10.1118/1.2761565.
Chen S, Zhou S, Zhang J, Marks L, Das S. WE‐D‐M100J‐04: Prediction of Lung Radiation‐Induced Pneumonitis Using the Support Vector Machine Algorithm. Medical Physics. 2007 Jan 1;34(6):2602–3.
Chen, S., et al. “WE‐D‐M100J‐04: Prediction of Lung Radiation‐Induced Pneumonitis Using the Support Vector Machine Algorithm.” Medical Physics, vol. 34, no. 6, Jan. 2007, pp. 2602–03. Scopus, doi:10.1118/1.2761565.
Chen S, Zhou S, Zhang J, Marks L, Das S. WE‐D‐M100J‐04: Prediction of Lung Radiation‐Induced Pneumonitis Using the Support Vector Machine Algorithm. Medical Physics. 2007 Jan 1;34(6):2602–2603.

Published In

Medical Physics

DOI

ISSN

0094-2405

Publication Date

January 1, 2007

Volume

34

Issue

6

Start / End Page

2602 / 2603

Related Subject Headings

  • Nuclear Medicine & Medical Imaging
  • 1112 Oncology and Carcinogenesis
  • 0903 Biomedical Engineering
  • 0299 Other Physical Sciences