Computer-Aided Diagnosis of Soft-Tissue Tumors Using Sonographic Morphologic and Texture Features

Chih Yen Chen, Hong Jen Chiou, Szu-Yuan Chou, See Ying Chiou, Hsin Kai Wang, Yi Hong Chou, Huihua Kenny Chiang

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Rationale and Objectives: The aim of this study was to develop a computer-aided diagnosis (CAD) system in assessing the sonographic morphologic and texture features of soft-tissue tumors. Materials and Methods: The retrospective study involved 114 pathology proven cases including 73 benign and 41 malignant soft-tissue tumors. The tumor regions were delineated by an experienced radiologist who was unknown to the pathologic result. Then, we applied 10 morphologic features and 6 gray-level co-occurrence matrix texture features to analyze the tumor regions. To classify the tumors as benign or malignant, we used two methods, a linear discriminant analysis with stepwise feature selection and a multilayer neural network with the back-propagation algorithm as classifiers. The classification performances are evaluated by the area Az under the receiver operating characteristic. Furthermore, four radiologists provided malignancy grades for all tumors in the comparison of the CAD system. Results: In this analysis, the CAD system based on the combination of morphologic and texture feature sets can give the optimal CAD result by LDA with an accuracy of 89.5%, a sensitivity of 90.2%, a specificity of 89.0%, a positive predictive value (PPV) of 82.2%, negative predictive value (NPV) of 94.2%, and Az value of 0.96, and by the multilayer perception with an accuracy of 88.6%, a sensitivity of 90.2%, a specificity of 87.5%, a positive predictive value of 80.4%, negative predictive value of 94.2%, and Az value of 0.95. The Az values of the four radiologists were ranged between 0.74 and 0.86, and the optimal CAD results were shown the highest Az values than the four radiologists' rankings. Conclusions: This study has shown that performing the CAD system with both morphologic and texture features on sonography, can successfully distinguish between benign and malignant soft-tissue tumors. Moreover, it can also provide a second opinion for the tumor diagnosis and avert unnecessary biopsy.

Original languageEnglish
Pages (from-to)1531-1538
Number of pages8
JournalAcademic Radiology
Volume16
Issue number12
DOIs
Publication statusPublished - Dec 2009

Fingerprint

Neoplasms
Discriminant Analysis
ROC Curve
Ultrasonography
Referral and Consultation
Retrospective Studies
Pathology
Biopsy
Radiologists

Keywords

  • computer-aided diagnosis (CAD)
  • linear discriminant analysis (LDA)
  • morphologic feature
  • multilayer perception (MLP)
  • soft-tissue tumors
  • texture feature

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Cite this

Chen, C. Y., Chiou, H. J., Chou, S-Y., Chiou, S. Y., Wang, H. K., Chou, Y. H., & Chiang, H. K. (2009). Computer-Aided Diagnosis of Soft-Tissue Tumors Using Sonographic Morphologic and Texture Features. Academic Radiology, 16(12), 1531-1538. https://doi.org/10.1016/j.acra.2009.07.024

Computer-Aided Diagnosis of Soft-Tissue Tumors Using Sonographic Morphologic and Texture Features. / Chen, Chih Yen; Chiou, Hong Jen; Chou, Szu-Yuan; Chiou, See Ying; Wang, Hsin Kai; Chou, Yi Hong; Chiang, Huihua Kenny.

In: Academic Radiology, Vol. 16, No. 12, 12.2009, p. 1531-1538.

Research output: Contribution to journalArticle

Chen, CY, Chiou, HJ, Chou, S-Y, Chiou, SY, Wang, HK, Chou, YH & Chiang, HK 2009, 'Computer-Aided Diagnosis of Soft-Tissue Tumors Using Sonographic Morphologic and Texture Features', Academic Radiology, vol. 16, no. 12, pp. 1531-1538. https://doi.org/10.1016/j.acra.2009.07.024
Chen, Chih Yen ; Chiou, Hong Jen ; Chou, Szu-Yuan ; Chiou, See Ying ; Wang, Hsin Kai ; Chou, Yi Hong ; Chiang, Huihua Kenny. / Computer-Aided Diagnosis of Soft-Tissue Tumors Using Sonographic Morphologic and Texture Features. In: Academic Radiology. 2009 ; Vol. 16, No. 12. pp. 1531-1538.
@article{abdec297524344898fa4293d53b1d62f,
title = "Computer-Aided Diagnosis of Soft-Tissue Tumors Using Sonographic Morphologic and Texture Features",
abstract = "Rationale and Objectives: The aim of this study was to develop a computer-aided diagnosis (CAD) system in assessing the sonographic morphologic and texture features of soft-tissue tumors. Materials and Methods: The retrospective study involved 114 pathology proven cases including 73 benign and 41 malignant soft-tissue tumors. The tumor regions were delineated by an experienced radiologist who was unknown to the pathologic result. Then, we applied 10 morphologic features and 6 gray-level co-occurrence matrix texture features to analyze the tumor regions. To classify the tumors as benign or malignant, we used two methods, a linear discriminant analysis with stepwise feature selection and a multilayer neural network with the back-propagation algorithm as classifiers. The classification performances are evaluated by the area Az under the receiver operating characteristic. Furthermore, four radiologists provided malignancy grades for all tumors in the comparison of the CAD system. Results: In this analysis, the CAD system based on the combination of morphologic and texture feature sets can give the optimal CAD result by LDA with an accuracy of 89.5{\%}, a sensitivity of 90.2{\%}, a specificity of 89.0{\%}, a positive predictive value (PPV) of 82.2{\%}, negative predictive value (NPV) of 94.2{\%}, and Az value of 0.96, and by the multilayer perception with an accuracy of 88.6{\%}, a sensitivity of 90.2{\%}, a specificity of 87.5{\%}, a positive predictive value of 80.4{\%}, negative predictive value of 94.2{\%}, and Az value of 0.95. The Az values of the four radiologists were ranged between 0.74 and 0.86, and the optimal CAD results were shown the highest Az values than the four radiologists' rankings. Conclusions: This study has shown that performing the CAD system with both morphologic and texture features on sonography, can successfully distinguish between benign and malignant soft-tissue tumors. Moreover, it can also provide a second opinion for the tumor diagnosis and avert unnecessary biopsy.",
keywords = "computer-aided diagnosis (CAD), linear discriminant analysis (LDA), morphologic feature, multilayer perception (MLP), soft-tissue tumors, texture feature",
author = "Chen, {Chih Yen} and Chiou, {Hong Jen} and Szu-Yuan Chou and Chiou, {See Ying} and Wang, {Hsin Kai} and Chou, {Yi Hong} and Chiang, {Huihua Kenny}",
year = "2009",
month = "12",
doi = "10.1016/j.acra.2009.07.024",
language = "English",
volume = "16",
pages = "1531--1538",
journal = "Academic Radiology",
issn = "1076-6332",
publisher = "Elsevier USA",
number = "12",

}

TY - JOUR

T1 - Computer-Aided Diagnosis of Soft-Tissue Tumors Using Sonographic Morphologic and Texture Features

AU - Chen, Chih Yen

AU - Chiou, Hong Jen

AU - Chou, Szu-Yuan

AU - Chiou, See Ying

AU - Wang, Hsin Kai

AU - Chou, Yi Hong

AU - Chiang, Huihua Kenny

PY - 2009/12

Y1 - 2009/12

N2 - Rationale and Objectives: The aim of this study was to develop a computer-aided diagnosis (CAD) system in assessing the sonographic morphologic and texture features of soft-tissue tumors. Materials and Methods: The retrospective study involved 114 pathology proven cases including 73 benign and 41 malignant soft-tissue tumors. The tumor regions were delineated by an experienced radiologist who was unknown to the pathologic result. Then, we applied 10 morphologic features and 6 gray-level co-occurrence matrix texture features to analyze the tumor regions. To classify the tumors as benign or malignant, we used two methods, a linear discriminant analysis with stepwise feature selection and a multilayer neural network with the back-propagation algorithm as classifiers. The classification performances are evaluated by the area Az under the receiver operating characteristic. Furthermore, four radiologists provided malignancy grades for all tumors in the comparison of the CAD system. Results: In this analysis, the CAD system based on the combination of morphologic and texture feature sets can give the optimal CAD result by LDA with an accuracy of 89.5%, a sensitivity of 90.2%, a specificity of 89.0%, a positive predictive value (PPV) of 82.2%, negative predictive value (NPV) of 94.2%, and Az value of 0.96, and by the multilayer perception with an accuracy of 88.6%, a sensitivity of 90.2%, a specificity of 87.5%, a positive predictive value of 80.4%, negative predictive value of 94.2%, and Az value of 0.95. The Az values of the four radiologists were ranged between 0.74 and 0.86, and the optimal CAD results were shown the highest Az values than the four radiologists' rankings. Conclusions: This study has shown that performing the CAD system with both morphologic and texture features on sonography, can successfully distinguish between benign and malignant soft-tissue tumors. Moreover, it can also provide a second opinion for the tumor diagnosis and avert unnecessary biopsy.

AB - Rationale and Objectives: The aim of this study was to develop a computer-aided diagnosis (CAD) system in assessing the sonographic morphologic and texture features of soft-tissue tumors. Materials and Methods: The retrospective study involved 114 pathology proven cases including 73 benign and 41 malignant soft-tissue tumors. The tumor regions were delineated by an experienced radiologist who was unknown to the pathologic result. Then, we applied 10 morphologic features and 6 gray-level co-occurrence matrix texture features to analyze the tumor regions. To classify the tumors as benign or malignant, we used two methods, a linear discriminant analysis with stepwise feature selection and a multilayer neural network with the back-propagation algorithm as classifiers. The classification performances are evaluated by the area Az under the receiver operating characteristic. Furthermore, four radiologists provided malignancy grades for all tumors in the comparison of the CAD system. Results: In this analysis, the CAD system based on the combination of morphologic and texture feature sets can give the optimal CAD result by LDA with an accuracy of 89.5%, a sensitivity of 90.2%, a specificity of 89.0%, a positive predictive value (PPV) of 82.2%, negative predictive value (NPV) of 94.2%, and Az value of 0.96, and by the multilayer perception with an accuracy of 88.6%, a sensitivity of 90.2%, a specificity of 87.5%, a positive predictive value of 80.4%, negative predictive value of 94.2%, and Az value of 0.95. The Az values of the four radiologists were ranged between 0.74 and 0.86, and the optimal CAD results were shown the highest Az values than the four radiologists' rankings. Conclusions: This study has shown that performing the CAD system with both morphologic and texture features on sonography, can successfully distinguish between benign and malignant soft-tissue tumors. Moreover, it can also provide a second opinion for the tumor diagnosis and avert unnecessary biopsy.

KW - computer-aided diagnosis (CAD)

KW - linear discriminant analysis (LDA)

KW - morphologic feature

KW - multilayer perception (MLP)

KW - soft-tissue tumors

KW - texture feature

UR - http://www.scopus.com/inward/record.url?scp=70350464946&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70350464946&partnerID=8YFLogxK

U2 - 10.1016/j.acra.2009.07.024

DO - 10.1016/j.acra.2009.07.024

M3 - Article

C2 - 19896070

AN - SCOPUS:70350464946

VL - 16

SP - 1531

EP - 1538

JO - Academic Radiology

JF - Academic Radiology

SN - 1076-6332

IS - 12

ER -