Impact of different b-value combinations on radiomics features of apparent diffusion coefficient in cervical cancer

Yaoyao He, Yi Rong, Hao Chen, Zhaoxi Zhang, Jianfeng Qiu, Lili Zheng, Stanley H Benedict, Xiaohui Niu, Ning Pan, Yulin Liu, Zilong Yuan

Research output: Contribution to journalArticle

Abstract

Background: The impact of variable b-value combinations on apparent diffusion coefficient (ADC)-based radiomics features has not been fully addressed in literature. Purpose: To investigate the correlation between radiomics features extracted from ADC maps and various b-value combinations in cervical cancer. Material and Methods: Diffusion-weighted images (b-values: 0, 600, 800, and 1000 s/mm2) of 20 patients with cervical cancer were included. Tumors were identified with the largest transversal cross-section and manually segmented by radiologist. For each b-value combination, 92 radiomics features were extracted and coefficient of variance (CV) was used to evaluate the robustness of radiomics features with different b-value combinations. Features with CV > 5% were normalized by the mean feature variation across the group. Results: Out of a total of 92 radiomics features, 18 were classified as robust features with CV ≤5%. Among the rest (CV > 5%), 11, 23, and 40 features demonstrated 5%< CV ≤10%, 10%< CV ≤20%, and CV > 20%, respectively. A subset of features in each category (CV > 5%) showed strong correlation with the b-value combination variation, including 44% (7/16) features in gray level co-occurrence matrix, 62% (8/13) features in gray level dependence matrix, 64% (9/14) features in first order, 50% (8/16) features in gray level run length matrix, 57% (8/14) features in gray level size matrix, and 20% (1/5) features in neighborhood gray-tone difference matrix. Conclusions: Variations in b-value combinations demonstrated impact on radiomics features extracted from ADC maps for cervical cancer. The radiomics features with CV <5% can be considered as robust features and are recommended to be used in multicenter radiomics studies.

Original languageEnglish (US)
JournalActa Radiologica
DOIs
StateAccepted/In press - Jan 1 2019

Fingerprint

Uterine Cervical Neoplasms
Multicenter Studies
Neoplasms

Keywords

  • apparent diffusion coefficient
  • b-value combination
  • cervical cancer
  • radiomics

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging

Cite this

Impact of different b-value combinations on radiomics features of apparent diffusion coefficient in cervical cancer. / He, Yaoyao; Rong, Yi; Chen, Hao; Zhang, Zhaoxi; Qiu, Jianfeng; Zheng, Lili; Benedict, Stanley H; Niu, Xiaohui; Pan, Ning; Liu, Yulin; Yuan, Zilong.

In: Acta Radiologica, 01.01.2019.

Research output: Contribution to journalArticle

He, Yaoyao ; Rong, Yi ; Chen, Hao ; Zhang, Zhaoxi ; Qiu, Jianfeng ; Zheng, Lili ; Benedict, Stanley H ; Niu, Xiaohui ; Pan, Ning ; Liu, Yulin ; Yuan, Zilong. / Impact of different b-value combinations on radiomics features of apparent diffusion coefficient in cervical cancer. In: Acta Radiologica. 2019.
@article{25d7dd157d684d55926f9279441885b3,
title = "Impact of different b-value combinations on radiomics features of apparent diffusion coefficient in cervical cancer",
abstract = "Background: The impact of variable b-value combinations on apparent diffusion coefficient (ADC)-based radiomics features has not been fully addressed in literature. Purpose: To investigate the correlation between radiomics features extracted from ADC maps and various b-value combinations in cervical cancer. Material and Methods: Diffusion-weighted images (b-values: 0, 600, 800, and 1000 s/mm2) of 20 patients with cervical cancer were included. Tumors were identified with the largest transversal cross-section and manually segmented by radiologist. For each b-value combination, 92 radiomics features were extracted and coefficient of variance (CV) was used to evaluate the robustness of radiomics features with different b-value combinations. Features with CV > 5{\%} were normalized by the mean feature variation across the group. Results: Out of a total of 92 radiomics features, 18 were classified as robust features with CV ≤5{\%}. Among the rest (CV > 5{\%}), 11, 23, and 40 features demonstrated 5{\%}< CV ≤10{\%}, 10{\%}< CV ≤20{\%}, and CV > 20{\%}, respectively. A subset of features in each category (CV > 5{\%}) showed strong correlation with the b-value combination variation, including 44{\%} (7/16) features in gray level co-occurrence matrix, 62{\%} (8/13) features in gray level dependence matrix, 64{\%} (9/14) features in first order, 50{\%} (8/16) features in gray level run length matrix, 57{\%} (8/14) features in gray level size matrix, and 20{\%} (1/5) features in neighborhood gray-tone difference matrix. Conclusions: Variations in b-value combinations demonstrated impact on radiomics features extracted from ADC maps for cervical cancer. The radiomics features with CV <5{\%} can be considered as robust features and are recommended to be used in multicenter radiomics studies.",
keywords = "apparent diffusion coefficient, b-value combination, cervical cancer, radiomics",
author = "Yaoyao He and Yi Rong and Hao Chen and Zhaoxi Zhang and Jianfeng Qiu and Lili Zheng and Benedict, {Stanley H} and Xiaohui Niu and Ning Pan and Yulin Liu and Zilong Yuan",
year = "2019",
month = "1",
day = "1",
doi = "10.1177/0284185119870157",
language = "English (US)",
journal = "Acta Radiologica",
issn = "0284-1851",
publisher = "SAGE Publications Ltd",

}

TY - JOUR

T1 - Impact of different b-value combinations on radiomics features of apparent diffusion coefficient in cervical cancer

AU - He, Yaoyao

AU - Rong, Yi

AU - Chen, Hao

AU - Zhang, Zhaoxi

AU - Qiu, Jianfeng

AU - Zheng, Lili

AU - Benedict, Stanley H

AU - Niu, Xiaohui

AU - Pan, Ning

AU - Liu, Yulin

AU - Yuan, Zilong

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Background: The impact of variable b-value combinations on apparent diffusion coefficient (ADC)-based radiomics features has not been fully addressed in literature. Purpose: To investigate the correlation between radiomics features extracted from ADC maps and various b-value combinations in cervical cancer. Material and Methods: Diffusion-weighted images (b-values: 0, 600, 800, and 1000 s/mm2) of 20 patients with cervical cancer were included. Tumors were identified with the largest transversal cross-section and manually segmented by radiologist. For each b-value combination, 92 radiomics features were extracted and coefficient of variance (CV) was used to evaluate the robustness of radiomics features with different b-value combinations. Features with CV > 5% were normalized by the mean feature variation across the group. Results: Out of a total of 92 radiomics features, 18 were classified as robust features with CV ≤5%. Among the rest (CV > 5%), 11, 23, and 40 features demonstrated 5%< CV ≤10%, 10%< CV ≤20%, and CV > 20%, respectively. A subset of features in each category (CV > 5%) showed strong correlation with the b-value combination variation, including 44% (7/16) features in gray level co-occurrence matrix, 62% (8/13) features in gray level dependence matrix, 64% (9/14) features in first order, 50% (8/16) features in gray level run length matrix, 57% (8/14) features in gray level size matrix, and 20% (1/5) features in neighborhood gray-tone difference matrix. Conclusions: Variations in b-value combinations demonstrated impact on radiomics features extracted from ADC maps for cervical cancer. The radiomics features with CV <5% can be considered as robust features and are recommended to be used in multicenter radiomics studies.

AB - Background: The impact of variable b-value combinations on apparent diffusion coefficient (ADC)-based radiomics features has not been fully addressed in literature. Purpose: To investigate the correlation between radiomics features extracted from ADC maps and various b-value combinations in cervical cancer. Material and Methods: Diffusion-weighted images (b-values: 0, 600, 800, and 1000 s/mm2) of 20 patients with cervical cancer were included. Tumors were identified with the largest transversal cross-section and manually segmented by radiologist. For each b-value combination, 92 radiomics features were extracted and coefficient of variance (CV) was used to evaluate the robustness of radiomics features with different b-value combinations. Features with CV > 5% were normalized by the mean feature variation across the group. Results: Out of a total of 92 radiomics features, 18 were classified as robust features with CV ≤5%. Among the rest (CV > 5%), 11, 23, and 40 features demonstrated 5%< CV ≤10%, 10%< CV ≤20%, and CV > 20%, respectively. A subset of features in each category (CV > 5%) showed strong correlation with the b-value combination variation, including 44% (7/16) features in gray level co-occurrence matrix, 62% (8/13) features in gray level dependence matrix, 64% (9/14) features in first order, 50% (8/16) features in gray level run length matrix, 57% (8/14) features in gray level size matrix, and 20% (1/5) features in neighborhood gray-tone difference matrix. Conclusions: Variations in b-value combinations demonstrated impact on radiomics features extracted from ADC maps for cervical cancer. The radiomics features with CV <5% can be considered as robust features and are recommended to be used in multicenter radiomics studies.

KW - apparent diffusion coefficient

KW - b-value combination

KW - cervical cancer

KW - radiomics

UR - http://www.scopus.com/inward/record.url?scp=85072172969&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85072172969&partnerID=8YFLogxK

U2 - 10.1177/0284185119870157

DO - 10.1177/0284185119870157

M3 - Article

JO - Acta Radiologica

JF - Acta Radiologica

SN - 0284-1851

ER -