Digital image analysis of EUS images accurately differentiates pancreatic cancer from chronic pancreatitis and normal tissue

Ananya Das, Cuong C Nguyen, Feng Li, Baoxin Li

Research output: Contribution to journalArticle

36 Citations (Scopus)

Abstract

Background: Concomitant changes of chronic pancreatitis markedly degrade the performance of EUS in diagnosing pancreatic adenocarcinoma (PC). Digital image analysis (DIA) of the spatial distribution of pixels in a US image has been used as an effective approach to tissue characterization. Objective: We applied the techniques of DIA to EUS images of the pancreas to develop a classification model capable of differentiating pancreatic adenocarcinoma from non-neoplastic tissue. Design: Representative regions of interest were digitally selected in EUS images of 3 groups of patients with normal pancreas (group I), chronic pancreatitis (group II), and pancreatic adenocarcinoma (group III). Texture analyses were then performed by using image analysis software. Principal component analysis (PCA) was used for data reduction, and, later, a neural-network-based predictive model was built, trained, and validated. Setting: Tertiary academic medical center. Patients: Patients undergoing EUS of the pancreas. Results: A total of 110, 99, and 110 regions of interest in groups I, II, III, respectively, were available for analysis. For each region, a total of 256 statistical parameters were extracted. Eleven parameters were subsequently retained by PCA. A neural network model was built, trained by using these parameters as input variables for prediction of PC, and then validated in the remainder of the data set. This model was very accurate in classifying PC with an area under the receiver operating characteristic curve of 0.93. Limitation: Exploratory study with a small number of patients. Conclusions: DIA of EUS images is accurate in differentiating PC from chronic inflammation and normal tissue. With the potential availability of real-time application, DIA can develop into a useful clinical diagnostic tool in pancreatic diseases and in certain situations may obviate EUS-guided FNA.

Original languageEnglish (US)
Pages (from-to)861-867
Number of pages7
JournalGastrointestinal Endoscopy
Volume67
Issue number6
DOIs
StatePublished - May 2008

Fingerprint

Chronic Pancreatitis
Pancreatic Neoplasms
Adenocarcinoma
Pancreas
Principal Component Analysis
Endoscopic Ultrasound-Guided Fine Needle Aspiration
Pancreatic Diseases
Public Opinion
Spatial Analysis
Neural Networks (Computer)
ROC Curve
Software
Inflammation

ASJC Scopus subject areas

  • Gastroenterology

Cite this

Digital image analysis of EUS images accurately differentiates pancreatic cancer from chronic pancreatitis and normal tissue. / Das, Ananya; Nguyen, Cuong C; Li, Feng; Li, Baoxin.

In: Gastrointestinal Endoscopy, Vol. 67, No. 6, 05.2008, p. 861-867.

Research output: Contribution to journalArticle

@article{c076693d22054de7ab65b43c9ce704ee,
title = "Digital image analysis of EUS images accurately differentiates pancreatic cancer from chronic pancreatitis and normal tissue",
abstract = "Background: Concomitant changes of chronic pancreatitis markedly degrade the performance of EUS in diagnosing pancreatic adenocarcinoma (PC). Digital image analysis (DIA) of the spatial distribution of pixels in a US image has been used as an effective approach to tissue characterization. Objective: We applied the techniques of DIA to EUS images of the pancreas to develop a classification model capable of differentiating pancreatic adenocarcinoma from non-neoplastic tissue. Design: Representative regions of interest were digitally selected in EUS images of 3 groups of patients with normal pancreas (group I), chronic pancreatitis (group II), and pancreatic adenocarcinoma (group III). Texture analyses were then performed by using image analysis software. Principal component analysis (PCA) was used for data reduction, and, later, a neural-network-based predictive model was built, trained, and validated. Setting: Tertiary academic medical center. Patients: Patients undergoing EUS of the pancreas. Results: A total of 110, 99, and 110 regions of interest in groups I, II, III, respectively, were available for analysis. For each region, a total of 256 statistical parameters were extracted. Eleven parameters were subsequently retained by PCA. A neural network model was built, trained by using these parameters as input variables for prediction of PC, and then validated in the remainder of the data set. This model was very accurate in classifying PC with an area under the receiver operating characteristic curve of 0.93. Limitation: Exploratory study with a small number of patients. Conclusions: DIA of EUS images is accurate in differentiating PC from chronic inflammation and normal tissue. With the potential availability of real-time application, DIA can develop into a useful clinical diagnostic tool in pancreatic diseases and in certain situations may obviate EUS-guided FNA.",
author = "Ananya Das and Nguyen, {Cuong C} and Feng Li and Baoxin Li",
year = "2008",
month = "5",
doi = "10.1016/j.gie.2007.08.036",
language = "English (US)",
volume = "67",
pages = "861--867",
journal = "Gastrointestinal Endoscopy",
issn = "0016-5107",
publisher = "Mosby Inc.",
number = "6",

}

TY - JOUR

T1 - Digital image analysis of EUS images accurately differentiates pancreatic cancer from chronic pancreatitis and normal tissue

AU - Das, Ananya

AU - Nguyen, Cuong C

AU - Li, Feng

AU - Li, Baoxin

PY - 2008/5

Y1 - 2008/5

N2 - Background: Concomitant changes of chronic pancreatitis markedly degrade the performance of EUS in diagnosing pancreatic adenocarcinoma (PC). Digital image analysis (DIA) of the spatial distribution of pixels in a US image has been used as an effective approach to tissue characterization. Objective: We applied the techniques of DIA to EUS images of the pancreas to develop a classification model capable of differentiating pancreatic adenocarcinoma from non-neoplastic tissue. Design: Representative regions of interest were digitally selected in EUS images of 3 groups of patients with normal pancreas (group I), chronic pancreatitis (group II), and pancreatic adenocarcinoma (group III). Texture analyses were then performed by using image analysis software. Principal component analysis (PCA) was used for data reduction, and, later, a neural-network-based predictive model was built, trained, and validated. Setting: Tertiary academic medical center. Patients: Patients undergoing EUS of the pancreas. Results: A total of 110, 99, and 110 regions of interest in groups I, II, III, respectively, were available for analysis. For each region, a total of 256 statistical parameters were extracted. Eleven parameters were subsequently retained by PCA. A neural network model was built, trained by using these parameters as input variables for prediction of PC, and then validated in the remainder of the data set. This model was very accurate in classifying PC with an area under the receiver operating characteristic curve of 0.93. Limitation: Exploratory study with a small number of patients. Conclusions: DIA of EUS images is accurate in differentiating PC from chronic inflammation and normal tissue. With the potential availability of real-time application, DIA can develop into a useful clinical diagnostic tool in pancreatic diseases and in certain situations may obviate EUS-guided FNA.

AB - Background: Concomitant changes of chronic pancreatitis markedly degrade the performance of EUS in diagnosing pancreatic adenocarcinoma (PC). Digital image analysis (DIA) of the spatial distribution of pixels in a US image has been used as an effective approach to tissue characterization. Objective: We applied the techniques of DIA to EUS images of the pancreas to develop a classification model capable of differentiating pancreatic adenocarcinoma from non-neoplastic tissue. Design: Representative regions of interest were digitally selected in EUS images of 3 groups of patients with normal pancreas (group I), chronic pancreatitis (group II), and pancreatic adenocarcinoma (group III). Texture analyses were then performed by using image analysis software. Principal component analysis (PCA) was used for data reduction, and, later, a neural-network-based predictive model was built, trained, and validated. Setting: Tertiary academic medical center. Patients: Patients undergoing EUS of the pancreas. Results: A total of 110, 99, and 110 regions of interest in groups I, II, III, respectively, were available for analysis. For each region, a total of 256 statistical parameters were extracted. Eleven parameters were subsequently retained by PCA. A neural network model was built, trained by using these parameters as input variables for prediction of PC, and then validated in the remainder of the data set. This model was very accurate in classifying PC with an area under the receiver operating characteristic curve of 0.93. Limitation: Exploratory study with a small number of patients. Conclusions: DIA of EUS images is accurate in differentiating PC from chronic inflammation and normal tissue. With the potential availability of real-time application, DIA can develop into a useful clinical diagnostic tool in pancreatic diseases and in certain situations may obviate EUS-guided FNA.

UR - http://www.scopus.com/inward/record.url?scp=42949167501&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=42949167501&partnerID=8YFLogxK

U2 - 10.1016/j.gie.2007.08.036

DO - 10.1016/j.gie.2007.08.036

M3 - Article

C2 - 18179797

AN - SCOPUS:42949167501

VL - 67

SP - 861

EP - 867

JO - Gastrointestinal Endoscopy

JF - Gastrointestinal Endoscopy

SN - 0016-5107

IS - 6

ER -