Comparative performance of random forest and support vector machine classifiers for detection of colorectal lesions in CT colonography

Janne J. Näppi, Daniele Regge, Hiroyuki Yoshida

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

A major problem of computer-aided detection (CAD) for computed tomographic colonography (CTC) is that CAD systems display large numbers of false-positive detections, thereby distracting users. Support vector machine (SVM) classifiers have been a popular choice for reducing false-positive detections in CAD systems. Recently, random forests (RF) have emerged as a novel type of highly accurate classifier. We compared the relative performance of RF and SVM classifiers in automated detection of colorectal lesions in CTC. The CAD system was trained with the CTC data of 123 patients and tested with an independent set of 737 patients. The results indicate that the performance of an RF classifier compares favorably with that of an SVM classifier in CTC.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages27-34
Number of pages8
Volume7029 LNCS
DOIs
Publication statusPublished - 2012
Event3rd International Workshop on Computational and Clinical Applications in Abdominal Imaging, Held in Conjunction with the 14th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2011 - Toronto, ON, Canada
Duration: Sep 18 2011Sep 18 2011

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7029 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other3rd International Workshop on Computational and Clinical Applications in Abdominal Imaging, Held in Conjunction with the 14th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2011
CountryCanada
CityToronto, ON
Period9/18/119/18/11

Fingerprint

Random Forest
Computer-aided Detection
Support vector machines
Support Vector Machine
Classifiers
Classifier
False Positive
Independent Set

Keywords

  • CAD
  • computer-aided detection
  • flat lesions
  • random forest
  • support vector machine
  • virtual colonoscopy
  • x-ray tomography

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Näppi, J. J., Regge, D., & Yoshida, H. (2012). Comparative performance of random forest and support vector machine classifiers for detection of colorectal lesions in CT colonography. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7029 LNCS, pp. 27-34). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7029 LNCS). https://doi.org/10.1007/978-3-642-28557-8_4

Comparative performance of random forest and support vector machine classifiers for detection of colorectal lesions in CT colonography. / Näppi, Janne J.; Regge, Daniele; Yoshida, Hiroyuki.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7029 LNCS 2012. p. 27-34 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7029 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Näppi, JJ, Regge, D & Yoshida, H 2012, Comparative performance of random forest and support vector machine classifiers for detection of colorectal lesions in CT colonography. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 7029 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7029 LNCS, pp. 27-34, 3rd International Workshop on Computational and Clinical Applications in Abdominal Imaging, Held in Conjunction with the 14th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2011, Toronto, ON, Canada, 9/18/11. https://doi.org/10.1007/978-3-642-28557-8_4
Näppi JJ, Regge D, Yoshida H. Comparative performance of random forest and support vector machine classifiers for detection of colorectal lesions in CT colonography. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7029 LNCS. 2012. p. 27-34. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-28557-8_4
Näppi, Janne J. ; Regge, Daniele ; Yoshida, Hiroyuki. / Comparative performance of random forest and support vector machine classifiers for detection of colorectal lesions in CT colonography. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7029 LNCS 2012. pp. 27-34 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{2a3f054d9e6340fda7c00722fb2e9c77,
title = "Comparative performance of random forest and support vector machine classifiers for detection of colorectal lesions in CT colonography",
abstract = "A major problem of computer-aided detection (CAD) for computed tomographic colonography (CTC) is that CAD systems display large numbers of false-positive detections, thereby distracting users. Support vector machine (SVM) classifiers have been a popular choice for reducing false-positive detections in CAD systems. Recently, random forests (RF) have emerged as a novel type of highly accurate classifier. We compared the relative performance of RF and SVM classifiers in automated detection of colorectal lesions in CTC. The CAD system was trained with the CTC data of 123 patients and tested with an independent set of 737 patients. The results indicate that the performance of an RF classifier compares favorably with that of an SVM classifier in CTC.",
keywords = "CAD, computer-aided detection, flat lesions, random forest, support vector machine, virtual colonoscopy, x-ray tomography",
author = "N{\"a}ppi, {Janne J.} and Daniele Regge and Hiroyuki Yoshida",
year = "2012",
doi = "10.1007/978-3-642-28557-8_4",
language = "English",
isbn = "9783642285561",
volume = "7029 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "27--34",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Comparative performance of random forest and support vector machine classifiers for detection of colorectal lesions in CT colonography

AU - Näppi, Janne J.

AU - Regge, Daniele

AU - Yoshida, Hiroyuki

PY - 2012

Y1 - 2012

N2 - A major problem of computer-aided detection (CAD) for computed tomographic colonography (CTC) is that CAD systems display large numbers of false-positive detections, thereby distracting users. Support vector machine (SVM) classifiers have been a popular choice for reducing false-positive detections in CAD systems. Recently, random forests (RF) have emerged as a novel type of highly accurate classifier. We compared the relative performance of RF and SVM classifiers in automated detection of colorectal lesions in CTC. The CAD system was trained with the CTC data of 123 patients and tested with an independent set of 737 patients. The results indicate that the performance of an RF classifier compares favorably with that of an SVM classifier in CTC.

AB - A major problem of computer-aided detection (CAD) for computed tomographic colonography (CTC) is that CAD systems display large numbers of false-positive detections, thereby distracting users. Support vector machine (SVM) classifiers have been a popular choice for reducing false-positive detections in CAD systems. Recently, random forests (RF) have emerged as a novel type of highly accurate classifier. We compared the relative performance of RF and SVM classifiers in automated detection of colorectal lesions in CTC. The CAD system was trained with the CTC data of 123 patients and tested with an independent set of 737 patients. The results indicate that the performance of an RF classifier compares favorably with that of an SVM classifier in CTC.

KW - CAD

KW - computer-aided detection

KW - flat lesions

KW - random forest

KW - support vector machine

KW - virtual colonoscopy

KW - x-ray tomography

UR - http://www.scopus.com/inward/record.url?scp=84858305498&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84858305498&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-28557-8_4

DO - 10.1007/978-3-642-28557-8_4

M3 - Conference contribution

AN - SCOPUS:84858305498

SN - 9783642285561

VL - 7029 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 27

EP - 34

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -