Share Email Print
cover

Proceedings Paper

Quantitative imaging features to predict cancer status in lung nodules
Author(s): Ying Liu; Yoganand Balagurunathan; Thomas Atwater; Sanja Antic; Qian Li; Ronald Walker; Gary T. Smith; Pierre P. Massion; Matthew B. Schabath; Robert J. Gillies
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Background: We propose a systematic methodology to quantify incidentally identified lung nodules based on observed radiological traits on a point scale. These quantitative traits classification model was used to predict cancer status. Materials and Methods: We used 102 patients’ low dose computed tomography (LDCT) images for this study, 24 semantic traits were systematically scored from each image. We built a machine learning classifier in cross validation setting to find best predictive imaging features to differentiate malignant from benign lung nodules. Results: The best feature triplet to discriminate malignancy was based on long axis, concavity and lymphadenopathy with average AUC of 0.897 (Accuracy of 76.8%, Sensitivity of 64.3%, Specificity of 90%). A similar semantic triplet optimized on Sensitivity/Specificity (Youden’s J index) included long axis, vascular convergence and lymphadenopathy which had an average AUC of 0.875 (Accuracy of 81.7%, Sensitivity of 76.2%, Specificity of 95%). Conclusions: Quantitative radiological image traits can differentiate malignant from benign lung nodules. These semantic features along with size measurement enhance the prediction accuracy.

Paper Details

Date Published: 24 March 2016
PDF: 6 pages
Proc. SPIE 9787, Medical Imaging 2016: Image Perception, Observer Performance, and Technology Assessment, 97870L (24 March 2016); doi: 10.1117/12.2216809
Show Author Affiliations
Ying Liu, Tianjin Medical Univ. Cancer Institute and Hospital (China)
H. Lee Moffitt Cancer Ctr. and Research Institute (United States)
Yoganand Balagurunathan, H. Lee Moffitt Cancer Ctr. and Research Institute (United States)
Thomas Atwater, Vanderbilt Univ. School of Medicine (United States)
Sanja Antic, Vanderbilt Univ. School of Medicine (United States)
Qian Li, Tianjin Medical Univ. Cancer Institute and Hospital (China)
H. Lee Moffitt Cancer Ctr. and Research Institute (United States)
Ronald Walker, Vanderbilt Univ. School of Medicine (United States)
Veterans Affairs Medical Ctr. (United States)
Gary T. Smith, Vanderbilt Univ. School of Medicine (United States)
Veterans Affairs Medical Ctr. (United States)
Pierre P. Massion, Vanderbilt Univ. School of Medicine (United States)
Veterans Affairs Medical Ctr. (United States)
Matthew B. Schabath, H. Lee Moffitt Cancer Ctr. and Research Institute (United States)
Robert J. Gillies, H. Lee Moffitt Cancer Ctr. and Research Institute (United States)


Published in SPIE Proceedings Vol. 9787:
Medical Imaging 2016: Image Perception, Observer Performance, and Technology Assessment
Craig K. Abbey; Matthew A. Kupinski, Editor(s)

© SPIE. Terms of Use
Back to Top