Automated polyp detection in colonoscopy videos using shape and context information

Nima Tajbakhsh, Suryakanth R. Gurudu, Jianming Liang

Research output: Contribution to journalArticle

85 Citations (Scopus)

Abstract

This paper presents the culmination of our research in designing a system for computer-aided detection (CAD) of polyps in colonoscopy videos. Our system is based on a hybrid context-shape approach, which utilizes context information to remove non-polyp structures and shape information to reliably localize polyps. Specifically, given a colonoscopy image, we first obtain a crude edge map. Second, we remove non-polyp edges from the edge map using our unique feature extraction and edge classification scheme. Third, we localize polyp candidates with probabilistic confidence scores in the refined edge maps using our novel voting scheme. The suggested CAD system has been tested using two public polyp databases, CVC-ColonDB, containing 300 colonoscopy images with a total of 300 polyp instances from 15 unique polyps, and ASU-Mayo database, which is our collection of colonoscopy videos containing 19,400 frames and a total of 5,200 polyp instances from 10 unique polyps. We have evaluated our system using free-response receiver operating characteristic (FROC) analysis. At 0.1 false positives per frame, our system achieves a sensitivity of 88.0% for CVC-ColonDB and a sensitivity of 48% for the ASU-Mayo database. In addition, we have evaluated our system using a new detection latency analysis where latency is defined as the time from the first appearance of a polyp in the colonoscopy video to the time of its first detection by our system. At 0.05 false positives per frame, our system yields a polyp detection latency of 0.3 seconds.

Original languageEnglish (US)
Article number2487997
Pages (from-to)630-644
Number of pages15
JournalIEEE Transactions on Medical Imaging
Volume35
Issue number2
DOIs
StatePublished - Feb 1 2016

Fingerprint

Colonoscopy
Polyps
Feature extraction
Databases
Computer Systems
Politics
ROC Curve

Keywords

  • Boundary classification
  • Detection latency
  • Edge voting
  • Optical colonoscopy
  • Polyp detection

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Science Applications
  • Radiological and Ultrasound Technology
  • Software

Cite this

Automated polyp detection in colonoscopy videos using shape and context information. / Tajbakhsh, Nima; Gurudu, Suryakanth R.; Liang, Jianming.

In: IEEE Transactions on Medical Imaging, Vol. 35, No. 2, 2487997, 01.02.2016, p. 630-644.

Research output: Contribution to journalArticle

Tajbakhsh, Nima ; Gurudu, Suryakanth R. ; Liang, Jianming. / Automated polyp detection in colonoscopy videos using shape and context information. In: IEEE Transactions on Medical Imaging. 2016 ; Vol. 35, No. 2. pp. 630-644.
@article{b5d6393121dc4080b7393c64594642d0,
title = "Automated polyp detection in colonoscopy videos using shape and context information",
abstract = "This paper presents the culmination of our research in designing a system for computer-aided detection (CAD) of polyps in colonoscopy videos. Our system is based on a hybrid context-shape approach, which utilizes context information to remove non-polyp structures and shape information to reliably localize polyps. Specifically, given a colonoscopy image, we first obtain a crude edge map. Second, we remove non-polyp edges from the edge map using our unique feature extraction and edge classification scheme. Third, we localize polyp candidates with probabilistic confidence scores in the refined edge maps using our novel voting scheme. The suggested CAD system has been tested using two public polyp databases, CVC-ColonDB, containing 300 colonoscopy images with a total of 300 polyp instances from 15 unique polyps, and ASU-Mayo database, which is our collection of colonoscopy videos containing 19,400 frames and a total of 5,200 polyp instances from 10 unique polyps. We have evaluated our system using free-response receiver operating characteristic (FROC) analysis. At 0.1 false positives per frame, our system achieves a sensitivity of 88.0{\%} for CVC-ColonDB and a sensitivity of 48{\%} for the ASU-Mayo database. In addition, we have evaluated our system using a new detection latency analysis where latency is defined as the time from the first appearance of a polyp in the colonoscopy video to the time of its first detection by our system. At 0.05 false positives per frame, our system yields a polyp detection latency of 0.3 seconds.",
keywords = "Boundary classification, Detection latency, Edge voting, Optical colonoscopy, Polyp detection",
author = "Nima Tajbakhsh and Gurudu, {Suryakanth R.} and Jianming Liang",
year = "2016",
month = "2",
day = "1",
doi = "10.1109/TMI.2015.2487997",
language = "English (US)",
volume = "35",
pages = "630--644",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - Automated polyp detection in colonoscopy videos using shape and context information

AU - Tajbakhsh, Nima

AU - Gurudu, Suryakanth R.

AU - Liang, Jianming

PY - 2016/2/1

Y1 - 2016/2/1

N2 - This paper presents the culmination of our research in designing a system for computer-aided detection (CAD) of polyps in colonoscopy videos. Our system is based on a hybrid context-shape approach, which utilizes context information to remove non-polyp structures and shape information to reliably localize polyps. Specifically, given a colonoscopy image, we first obtain a crude edge map. Second, we remove non-polyp edges from the edge map using our unique feature extraction and edge classification scheme. Third, we localize polyp candidates with probabilistic confidence scores in the refined edge maps using our novel voting scheme. The suggested CAD system has been tested using two public polyp databases, CVC-ColonDB, containing 300 colonoscopy images with a total of 300 polyp instances from 15 unique polyps, and ASU-Mayo database, which is our collection of colonoscopy videos containing 19,400 frames and a total of 5,200 polyp instances from 10 unique polyps. We have evaluated our system using free-response receiver operating characteristic (FROC) analysis. At 0.1 false positives per frame, our system achieves a sensitivity of 88.0% for CVC-ColonDB and a sensitivity of 48% for the ASU-Mayo database. In addition, we have evaluated our system using a new detection latency analysis where latency is defined as the time from the first appearance of a polyp in the colonoscopy video to the time of its first detection by our system. At 0.05 false positives per frame, our system yields a polyp detection latency of 0.3 seconds.

AB - This paper presents the culmination of our research in designing a system for computer-aided detection (CAD) of polyps in colonoscopy videos. Our system is based on a hybrid context-shape approach, which utilizes context information to remove non-polyp structures and shape information to reliably localize polyps. Specifically, given a colonoscopy image, we first obtain a crude edge map. Second, we remove non-polyp edges from the edge map using our unique feature extraction and edge classification scheme. Third, we localize polyp candidates with probabilistic confidence scores in the refined edge maps using our novel voting scheme. The suggested CAD system has been tested using two public polyp databases, CVC-ColonDB, containing 300 colonoscopy images with a total of 300 polyp instances from 15 unique polyps, and ASU-Mayo database, which is our collection of colonoscopy videos containing 19,400 frames and a total of 5,200 polyp instances from 10 unique polyps. We have evaluated our system using free-response receiver operating characteristic (FROC) analysis. At 0.1 false positives per frame, our system achieves a sensitivity of 88.0% for CVC-ColonDB and a sensitivity of 48% for the ASU-Mayo database. In addition, we have evaluated our system using a new detection latency analysis where latency is defined as the time from the first appearance of a polyp in the colonoscopy video to the time of its first detection by our system. At 0.05 false positives per frame, our system yields a polyp detection latency of 0.3 seconds.

KW - Boundary classification

KW - Detection latency

KW - Edge voting

KW - Optical colonoscopy

KW - Polyp detection

UR - http://www.scopus.com/inward/record.url?scp=84959449815&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959449815&partnerID=8YFLogxK

U2 - 10.1109/TMI.2015.2487997

DO - 10.1109/TMI.2015.2487997

M3 - Article

C2 - 26462083

AN - SCOPUS:84959449815

VL - 35

SP - 630

EP - 644

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 2

M1 - 2487997

ER -