Novel high-resolution computed tomography-based radiomic classifier for screen-identified pulmonary nodules in the National Lung Screening Trial

Tobias D Peikert, Fenghai Duan, Srinivasan Rajagopalan, Ronald A. Karwoski, Ryan Clay, Richard A. Robb, Ziling Qin, Jo Rean Sicks, Brian Jack Bartholmai, Fabien Maldonado

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Purpose Optimization of the clinical management of screen-detected lung nodules is needed to avoid unnecessary diagnostic interventions. Herein we demonstrate the potential value of a novel radiomics-based approach for the classification of screen-detected indeterminate nodules.=Material and methods Independent quantitative variables assessing various radiologic nodule features such as sphericity, flatness, elongation, spiculation, lobulation and curvature were developed from the NLST dataset using 726 indeterminate nodules (all 7 mm, benign, n = 318 and malignant, n = 408). Multivariate analysis was performed using least absolute shrinkage and selection operator (LASSO) method for variable selection and regularization in order to enhance the prediction accuracy and interpretability of the multivariate model. The bootstrapping method was then applied for the internal validation and the optimism-corrected AUC was reported for the final model.Results Eight of the originally considered 57 quantitative radiologic features were selected by LASSO multivariate modeling. These 8 features include variables capturing Location: vertical location (Offset carina centroid z), Size: volume estimate (Minimum enclosing brick), Shape: flatness, Density: texture analysis (Score Indicative of Lesion/Lung Aggression/ Abnormality (SILA) texture), and surface characteristics: surface complexity (Maximum shape index and Average shape index), and estimates of surface curvature (Average positive mean curvature and Minimum mean curvature), all with P<0.01. The optimism-corrected AUC for these 8 features is 0.939.Conclusions Our novel radiomic LDCT-based approach for indeterminate screen-detected nodule characterization appears extremely promising however independent external validation is needed.

Original languageEnglish (US)
Article numbere0196910
JournalPLoS One
Volume13
Issue number5
DOIs
StatePublished - May 1 2018

Fingerprint

computed tomography
Tomography
Screening
Classifiers
lungs
screening
shrinkage
Lung
Area Under Curve
Textures
texture
bricks
Brick
Aggression
lesions (animal)
multivariate analysis
aggression
Elongation
Multivariate Analysis
methodology

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)

Cite this

Novel high-resolution computed tomography-based radiomic classifier for screen-identified pulmonary nodules in the National Lung Screening Trial. / Peikert, Tobias D; Duan, Fenghai; Rajagopalan, Srinivasan; Karwoski, Ronald A.; Clay, Ryan; Robb, Richard A.; Qin, Ziling; Sicks, Jo Rean; Bartholmai, Brian Jack; Maldonado, Fabien.

In: PLoS One, Vol. 13, No. 5, e0196910, 01.05.2018.

Research output: Contribution to journalArticle

Peikert, Tobias D ; Duan, Fenghai ; Rajagopalan, Srinivasan ; Karwoski, Ronald A. ; Clay, Ryan ; Robb, Richard A. ; Qin, Ziling ; Sicks, Jo Rean ; Bartholmai, Brian Jack ; Maldonado, Fabien. / Novel high-resolution computed tomography-based radiomic classifier for screen-identified pulmonary nodules in the National Lung Screening Trial. In: PLoS One. 2018 ; Vol. 13, No. 5.
@article{77d56c3e22ff4952a8130388a61cf4b7,
title = "Novel high-resolution computed tomography-based radiomic classifier for screen-identified pulmonary nodules in the National Lung Screening Trial",
abstract = "Purpose Optimization of the clinical management of screen-detected lung nodules is needed to avoid unnecessary diagnostic interventions. Herein we demonstrate the potential value of a novel radiomics-based approach for the classification of screen-detected indeterminate nodules.=Material and methods Independent quantitative variables assessing various radiologic nodule features such as sphericity, flatness, elongation, spiculation, lobulation and curvature were developed from the NLST dataset using 726 indeterminate nodules (all 7 mm, benign, n = 318 and malignant, n = 408). Multivariate analysis was performed using least absolute shrinkage and selection operator (LASSO) method for variable selection and regularization in order to enhance the prediction accuracy and interpretability of the multivariate model. The bootstrapping method was then applied for the internal validation and the optimism-corrected AUC was reported for the final model.Results Eight of the originally considered 57 quantitative radiologic features were selected by LASSO multivariate modeling. These 8 features include variables capturing Location: vertical location (Offset carina centroid z), Size: volume estimate (Minimum enclosing brick), Shape: flatness, Density: texture analysis (Score Indicative of Lesion/Lung Aggression/ Abnormality (SILA) texture), and surface characteristics: surface complexity (Maximum shape index and Average shape index), and estimates of surface curvature (Average positive mean curvature and Minimum mean curvature), all with P<0.01. The optimism-corrected AUC for these 8 features is 0.939.Conclusions Our novel radiomic LDCT-based approach for indeterminate screen-detected nodule characterization appears extremely promising however independent external validation is needed.",
author = "Peikert, {Tobias D} and Fenghai Duan and Srinivasan Rajagopalan and Karwoski, {Ronald A.} and Ryan Clay and Robb, {Richard A.} and Ziling Qin and Sicks, {Jo Rean} and Bartholmai, {Brian Jack} and Fabien Maldonado",
year = "2018",
month = "5",
day = "1",
doi = "10.1371/journal.pone.0196910",
language = "English (US)",
volume = "13",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "5",

}

TY - JOUR

T1 - Novel high-resolution computed tomography-based radiomic classifier for screen-identified pulmonary nodules in the National Lung Screening Trial

AU - Peikert, Tobias D

AU - Duan, Fenghai

AU - Rajagopalan, Srinivasan

AU - Karwoski, Ronald A.

AU - Clay, Ryan

AU - Robb, Richard A.

AU - Qin, Ziling

AU - Sicks, Jo Rean

AU - Bartholmai, Brian Jack

AU - Maldonado, Fabien

PY - 2018/5/1

Y1 - 2018/5/1

N2 - Purpose Optimization of the clinical management of screen-detected lung nodules is needed to avoid unnecessary diagnostic interventions. Herein we demonstrate the potential value of a novel radiomics-based approach for the classification of screen-detected indeterminate nodules.=Material and methods Independent quantitative variables assessing various radiologic nodule features such as sphericity, flatness, elongation, spiculation, lobulation and curvature were developed from the NLST dataset using 726 indeterminate nodules (all 7 mm, benign, n = 318 and malignant, n = 408). Multivariate analysis was performed using least absolute shrinkage and selection operator (LASSO) method for variable selection and regularization in order to enhance the prediction accuracy and interpretability of the multivariate model. The bootstrapping method was then applied for the internal validation and the optimism-corrected AUC was reported for the final model.Results Eight of the originally considered 57 quantitative radiologic features were selected by LASSO multivariate modeling. These 8 features include variables capturing Location: vertical location (Offset carina centroid z), Size: volume estimate (Minimum enclosing brick), Shape: flatness, Density: texture analysis (Score Indicative of Lesion/Lung Aggression/ Abnormality (SILA) texture), and surface characteristics: surface complexity (Maximum shape index and Average shape index), and estimates of surface curvature (Average positive mean curvature and Minimum mean curvature), all with P<0.01. The optimism-corrected AUC for these 8 features is 0.939.Conclusions Our novel radiomic LDCT-based approach for indeterminate screen-detected nodule characterization appears extremely promising however independent external validation is needed.

AB - Purpose Optimization of the clinical management of screen-detected lung nodules is needed to avoid unnecessary diagnostic interventions. Herein we demonstrate the potential value of a novel radiomics-based approach for the classification of screen-detected indeterminate nodules.=Material and methods Independent quantitative variables assessing various radiologic nodule features such as sphericity, flatness, elongation, spiculation, lobulation and curvature were developed from the NLST dataset using 726 indeterminate nodules (all 7 mm, benign, n = 318 and malignant, n = 408). Multivariate analysis was performed using least absolute shrinkage and selection operator (LASSO) method for variable selection and regularization in order to enhance the prediction accuracy and interpretability of the multivariate model. The bootstrapping method was then applied for the internal validation and the optimism-corrected AUC was reported for the final model.Results Eight of the originally considered 57 quantitative radiologic features were selected by LASSO multivariate modeling. These 8 features include variables capturing Location: vertical location (Offset carina centroid z), Size: volume estimate (Minimum enclosing brick), Shape: flatness, Density: texture analysis (Score Indicative of Lesion/Lung Aggression/ Abnormality (SILA) texture), and surface characteristics: surface complexity (Maximum shape index and Average shape index), and estimates of surface curvature (Average positive mean curvature and Minimum mean curvature), all with P<0.01. The optimism-corrected AUC for these 8 features is 0.939.Conclusions Our novel radiomic LDCT-based approach for indeterminate screen-detected nodule characterization appears extremely promising however independent external validation is needed.

UR - http://www.scopus.com/inward/record.url?scp=85046964170&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046964170&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0196910

DO - 10.1371/journal.pone.0196910

M3 - Article

C2 - 29758038

AN - SCOPUS:85046964170

VL - 13

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 5

M1 - e0196910

ER -