Nonlinear kernel-based approaches for predicting normal tissue toxicities

Issam El Naqa, Jeffrey D. Bradley, Joseph O. Deasy

Research output: Chapter in Book/Report/Conference proceedingConference contribution

10 Scopus citations

Abstract

Since the early demonstration of the curative potential of radiation therapy for tumor sterilization, normal tissue toxicity continues to be dose limiting. Accurate prediction of patient's complication risk would allow personalization of treatment planning decisions. Nonlinear kernel methods can provide a robust framework for learning complex interactions between observed toxicities and treatment, anatomical, and patient-related variables. However, proper application of these powerful methods would require better understanding of a high-dimensional feature space that is spanned by all these variables. In this work, we investigate methods for visualization of this high-dimensional space and compare different approaches for extracting discriminant features. Our preliminary results demonstrate that principle component analysis is a valuable tool for visualizing high dimensional data and for determining proper kernel type. In addition, variable selection based on resampling methods within the logistic regression framework seemed to yield improved prediction performance compared to the recursive-feature elimination method.

Original languageEnglish
Title of host publicationProceedings - 7th International Conference on Machine Learning and Applications, ICMLA 2008
Pages539-544
Number of pages6
DOIs
StatePublished - Dec 1 2008
Event7th International Conference on Machine Learning and Applications, ICMLA 2008 - San Diego, CA, United States
Duration: Dec 11 2008Dec 13 2008

Publication series

NameProceedings - 7th International Conference on Machine Learning and Applications, ICMLA 2008

Conference

Conference7th International Conference on Machine Learning and Applications, ICMLA 2008
CountryUnited States
CitySan Diego, CA
Period12/11/0812/13/08

Fingerprint Dive into the research topics of 'Nonlinear kernel-based approaches for predicting normal tissue toxicities'. Together they form a unique fingerprint.

  • Cite this

    El Naqa, I., Bradley, J. D., & Deasy, J. O. (2008). Nonlinear kernel-based approaches for predicting normal tissue toxicities. In Proceedings - 7th International Conference on Machine Learning and Applications, ICMLA 2008 (pp. 539-544). [4725026] (Proceedings - 7th International Conference on Machine Learning and Applications, ICMLA 2008). https://doi.org/10.1109/ICMLA.2008.126