Digital image analysis is a useful adjunct to endoscopic ultrasonographic diagnosis of subepithelial lesions of the gastrointestinal tract

Vien X. Nguyen, Cuong C Nguyen, Baoxin Li, Ananya Das

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Objective: The purpose of this study was to explore the role of digital image analysis in differentiating endoscopic ultrasonographic (EUS) features of potentially malignant gastrointestinal subepithelial lesions (SELs) from those of benign lesions. Methods: Forty-six patients with histopathologically confirmed gastrointestinal stromal tumors (GISTs), carcinoids, and lipomas who had undergone EUS evaluation were identified from our database. Representative regions of interest (ROIs) were selected from the EUS images, and features were extracted by texture analysis. On the basis of these features, an artificial neural network (ANN) was built, trained, and internally validated by unsupervised learning followed by supervised learning. Outcomes were the performance characteristics of the ANN. Results: A total of 106, 111, and 124 ROIs were selected from EUS images of 8, 10, and 28 patients with lipomas, carcinoids, and GISTs, respectively. For each ROI, 228 statistical parameters were extracted and later reduced to the 11 most informative features by principal component analysis. After training with 50% of the data, the remainder of the data were used to validate the ANN. The model was "good" in differentiating carcinoids and GISTs, with area under the receiver operating characteristic curve (AUC) values of 0.86 and 0.89, respectively. The model was "excellent" in identifying lipomas correctly, with an AUC of 0.92. Conclusions: Digital image analysis of EUS images is a useful noninvasive adjunct to EUS evaluation of SELs.

Original languageEnglish (US)
Pages (from-to)1345-1351
Number of pages7
JournalJournal of Ultrasound in Medicine
Volume29
Issue number9
StatePublished - Sep 1 2010

Fingerprint

Gastrointestinal Stromal Tumors
Lipoma
Carcinoid Tumor
Gastrointestinal Tract
Area Under Curve
Learning
Principal Component Analysis
ROC Curve
Databases

Keywords

  • Artificial neural network
  • Digital image analysis
  • Endoscopic ultrasonography
  • Subepithelial lesions
  • Submucosal lesions

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Radiological and Ultrasound Technology
  • Medicine(all)

Cite this

Digital image analysis is a useful adjunct to endoscopic ultrasonographic diagnosis of subepithelial lesions of the gastrointestinal tract. / Nguyen, Vien X.; Nguyen, Cuong C; Li, Baoxin; Das, Ananya.

In: Journal of Ultrasound in Medicine, Vol. 29, No. 9, 01.09.2010, p. 1345-1351.

Research output: Contribution to journalArticle

@article{22869d438f1b41f5be84c7f9a45f79ef,
title = "Digital image analysis is a useful adjunct to endoscopic ultrasonographic diagnosis of subepithelial lesions of the gastrointestinal tract",
abstract = "Objective: The purpose of this study was to explore the role of digital image analysis in differentiating endoscopic ultrasonographic (EUS) features of potentially malignant gastrointestinal subepithelial lesions (SELs) from those of benign lesions. Methods: Forty-six patients with histopathologically confirmed gastrointestinal stromal tumors (GISTs), carcinoids, and lipomas who had undergone EUS evaluation were identified from our database. Representative regions of interest (ROIs) were selected from the EUS images, and features were extracted by texture analysis. On the basis of these features, an artificial neural network (ANN) was built, trained, and internally validated by unsupervised learning followed by supervised learning. Outcomes were the performance characteristics of the ANN. Results: A total of 106, 111, and 124 ROIs were selected from EUS images of 8, 10, and 28 patients with lipomas, carcinoids, and GISTs, respectively. For each ROI, 228 statistical parameters were extracted and later reduced to the 11 most informative features by principal component analysis. After training with 50{\%} of the data, the remainder of the data were used to validate the ANN. The model was {"}good{"} in differentiating carcinoids and GISTs, with area under the receiver operating characteristic curve (AUC) values of 0.86 and 0.89, respectively. The model was {"}excellent{"} in identifying lipomas correctly, with an AUC of 0.92. Conclusions: Digital image analysis of EUS images is a useful noninvasive adjunct to EUS evaluation of SELs.",
keywords = "Artificial neural network, Digital image analysis, Endoscopic ultrasonography, Subepithelial lesions, Submucosal lesions",
author = "Nguyen, {Vien X.} and Nguyen, {Cuong C} and Baoxin Li and Ananya Das",
year = "2010",
month = "9",
day = "1",
language = "English (US)",
volume = "29",
pages = "1345--1351",
journal = "Journal of Ultrasound in Medicine",
issn = "0278-4297",
publisher = "American Institute of Ultrasound in Medicine",
number = "9",

}

TY - JOUR

T1 - Digital image analysis is a useful adjunct to endoscopic ultrasonographic diagnosis of subepithelial lesions of the gastrointestinal tract

AU - Nguyen, Vien X.

AU - Nguyen, Cuong C

AU - Li, Baoxin

AU - Das, Ananya

PY - 2010/9/1

Y1 - 2010/9/1

N2 - Objective: The purpose of this study was to explore the role of digital image analysis in differentiating endoscopic ultrasonographic (EUS) features of potentially malignant gastrointestinal subepithelial lesions (SELs) from those of benign lesions. Methods: Forty-six patients with histopathologically confirmed gastrointestinal stromal tumors (GISTs), carcinoids, and lipomas who had undergone EUS evaluation were identified from our database. Representative regions of interest (ROIs) were selected from the EUS images, and features were extracted by texture analysis. On the basis of these features, an artificial neural network (ANN) was built, trained, and internally validated by unsupervised learning followed by supervised learning. Outcomes were the performance characteristics of the ANN. Results: A total of 106, 111, and 124 ROIs were selected from EUS images of 8, 10, and 28 patients with lipomas, carcinoids, and GISTs, respectively. For each ROI, 228 statistical parameters were extracted and later reduced to the 11 most informative features by principal component analysis. After training with 50% of the data, the remainder of the data were used to validate the ANN. The model was "good" in differentiating carcinoids and GISTs, with area under the receiver operating characteristic curve (AUC) values of 0.86 and 0.89, respectively. The model was "excellent" in identifying lipomas correctly, with an AUC of 0.92. Conclusions: Digital image analysis of EUS images is a useful noninvasive adjunct to EUS evaluation of SELs.

AB - Objective: The purpose of this study was to explore the role of digital image analysis in differentiating endoscopic ultrasonographic (EUS) features of potentially malignant gastrointestinal subepithelial lesions (SELs) from those of benign lesions. Methods: Forty-six patients with histopathologically confirmed gastrointestinal stromal tumors (GISTs), carcinoids, and lipomas who had undergone EUS evaluation were identified from our database. Representative regions of interest (ROIs) were selected from the EUS images, and features were extracted by texture analysis. On the basis of these features, an artificial neural network (ANN) was built, trained, and internally validated by unsupervised learning followed by supervised learning. Outcomes were the performance characteristics of the ANN. Results: A total of 106, 111, and 124 ROIs were selected from EUS images of 8, 10, and 28 patients with lipomas, carcinoids, and GISTs, respectively. For each ROI, 228 statistical parameters were extracted and later reduced to the 11 most informative features by principal component analysis. After training with 50% of the data, the remainder of the data were used to validate the ANN. The model was "good" in differentiating carcinoids and GISTs, with area under the receiver operating characteristic curve (AUC) values of 0.86 and 0.89, respectively. The model was "excellent" in identifying lipomas correctly, with an AUC of 0.92. Conclusions: Digital image analysis of EUS images is a useful noninvasive adjunct to EUS evaluation of SELs.

KW - Artificial neural network

KW - Digital image analysis

KW - Endoscopic ultrasonography

KW - Subepithelial lesions

KW - Submucosal lesions

UR - http://www.scopus.com/inward/record.url?scp=77956038850&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77956038850&partnerID=8YFLogxK

M3 - Article

C2 - 20733191

AN - SCOPUS:77956038850

VL - 29

SP - 1345

EP - 1351

JO - Journal of Ultrasound in Medicine

JF - Journal of Ultrasound in Medicine

SN - 0278-4297

IS - 9

ER -