CT colonography

Influence of 3D viewing and polyp candidate features on interpretation with computer-aided detection

Rong Shi, Pamela Schraedley-Desmond, Sandy Napel, Eric W. Olcott, R. Brooke Jeffrey, Judy Yee, Michael E. Zalis, Daniel Margolis, David S. Paik, Anthony J. Sherbondy, Padmavathi Sundaram, Christopher F. Beaulieu

Research output: Contribution to journalReview article

25 Citations (Scopus)

Abstract

Purpose: To retrospectively determine if three-dimensional (3D) viewing improves radiologists' accuracy in classifying true-positive (TP) and false-positive (FP) polyp candidates identified with computer-aided detection (CAD) and to determine candidate polyp features that are associated with classification accuracy, with known polyps serving as the reference standard. Materials and Methods: Institutional review board approval and informed consent were obtained; this study was HIPAA compliant. Forty-seven computed tomographic (CT) colonography data sets were obtained in 26 men and 10 women (age range, 42-76 years). Four radiologists classified 705 polyp candidates (53 TP candidates, 652 FP candidates) identified with CAD; initially, only two-dimensional images were used, but these were later supplemented with 3D rendering. Another radiologist unblinded to colonoscopy findings characterized the features of each candidate, assessed co-Ion distention and preparation, and defined the true nature of FP candidates. Receiver operating characteristic curves were used to compare readers' performance, and repeated-measures analysis of variance was used to test features that affect interpretation. Results: Use of 3D viewing improved classification accuracy for three readers and increased the area under the receiver operating characteristic curve to 0.96-0.97 (P < .001). For TP candidates, maximum polyp width (P = .038), polyp height (P = .019), and preparation (P = .004) significantly affected accuracy. For FP candidates, colonic segment (P = .007), attenuation (P < .001), surface smoothness (P < .001), distention (P = .034), preparation (P < .001), and true nature of candidate lesions (P < .001) significantly affected accuracy. Conclusion: Use of 3D viewing increases reader accuracy in the classification of polyp candidates identified with CAD. Polyp size and examination quality are significantly associated with accuracy.

Original languageEnglish (US)
Pages (from-to)768-776
Number of pages9
JournalRadiology
Volume239
Issue number3
DOIs
StatePublished - Jun 1 2006
Externally publishedYes

Fingerprint

Computed Tomographic Colonography
Polyps
ROC Curve
Health Insurance Portability and Accountability Act
Research Ethics Committees
Colonoscopy
Informed Consent
Analysis of Variance
Ions

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Cite this

Shi, R., Schraedley-Desmond, P., Napel, S., Olcott, E. W., Jeffrey, R. B., Yee, J., ... Beaulieu, C. F. (2006). CT colonography: Influence of 3D viewing and polyp candidate features on interpretation with computer-aided detection. Radiology, 239(3), 768-776. https://doi.org/10.1148/radiol.2393050418

CT colonography : Influence of 3D viewing and polyp candidate features on interpretation with computer-aided detection. / Shi, Rong; Schraedley-Desmond, Pamela; Napel, Sandy; Olcott, Eric W.; Jeffrey, R. Brooke; Yee, Judy; Zalis, Michael E.; Margolis, Daniel; Paik, David S.; Sherbondy, Anthony J.; Sundaram, Padmavathi; Beaulieu, Christopher F.

In: Radiology, Vol. 239, No. 3, 01.06.2006, p. 768-776.

Research output: Contribution to journalReview article

Shi, R, Schraedley-Desmond, P, Napel, S, Olcott, EW, Jeffrey, RB, Yee, J, Zalis, ME, Margolis, D, Paik, DS, Sherbondy, AJ, Sundaram, P & Beaulieu, CF 2006, 'CT colonography: Influence of 3D viewing and polyp candidate features on interpretation with computer-aided detection', Radiology, vol. 239, no. 3, pp. 768-776. https://doi.org/10.1148/radiol.2393050418
Shi, Rong ; Schraedley-Desmond, Pamela ; Napel, Sandy ; Olcott, Eric W. ; Jeffrey, R. Brooke ; Yee, Judy ; Zalis, Michael E. ; Margolis, Daniel ; Paik, David S. ; Sherbondy, Anthony J. ; Sundaram, Padmavathi ; Beaulieu, Christopher F. / CT colonography : Influence of 3D viewing and polyp candidate features on interpretation with computer-aided detection. In: Radiology. 2006 ; Vol. 239, No. 3. pp. 768-776.
@article{32760f96685943ea801ebacca0520f40,
title = "CT colonography: Influence of 3D viewing and polyp candidate features on interpretation with computer-aided detection",
abstract = "Purpose: To retrospectively determine if three-dimensional (3D) viewing improves radiologists' accuracy in classifying true-positive (TP) and false-positive (FP) polyp candidates identified with computer-aided detection (CAD) and to determine candidate polyp features that are associated with classification accuracy, with known polyps serving as the reference standard. Materials and Methods: Institutional review board approval and informed consent were obtained; this study was HIPAA compliant. Forty-seven computed tomographic (CT) colonography data sets were obtained in 26 men and 10 women (age range, 42-76 years). Four radiologists classified 705 polyp candidates (53 TP candidates, 652 FP candidates) identified with CAD; initially, only two-dimensional images were used, but these were later supplemented with 3D rendering. Another radiologist unblinded to colonoscopy findings characterized the features of each candidate, assessed co-Ion distention and preparation, and defined the true nature of FP candidates. Receiver operating characteristic curves were used to compare readers' performance, and repeated-measures analysis of variance was used to test features that affect interpretation. Results: Use of 3D viewing improved classification accuracy for three readers and increased the area under the receiver operating characteristic curve to 0.96-0.97 (P < .001). For TP candidates, maximum polyp width (P = .038), polyp height (P = .019), and preparation (P = .004) significantly affected accuracy. For FP candidates, colonic segment (P = .007), attenuation (P < .001), surface smoothness (P < .001), distention (P = .034), preparation (P < .001), and true nature of candidate lesions (P < .001) significantly affected accuracy. Conclusion: Use of 3D viewing increases reader accuracy in the classification of polyp candidates identified with CAD. Polyp size and examination quality are significantly associated with accuracy.",
author = "Rong Shi and Pamela Schraedley-Desmond and Sandy Napel and Olcott, {Eric W.} and Jeffrey, {R. Brooke} and Judy Yee and Zalis, {Michael E.} and Daniel Margolis and Paik, {David S.} and Sherbondy, {Anthony J.} and Padmavathi Sundaram and Beaulieu, {Christopher F.}",
year = "2006",
month = "6",
day = "1",
doi = "10.1148/radiol.2393050418",
language = "English (US)",
volume = "239",
pages = "768--776",
journal = "Radiology",
issn = "0033-8419",
publisher = "Radiological Society of North America Inc.",
number = "3",

}

TY - JOUR

T1 - CT colonography

T2 - Influence of 3D viewing and polyp candidate features on interpretation with computer-aided detection

AU - Shi, Rong

AU - Schraedley-Desmond, Pamela

AU - Napel, Sandy

AU - Olcott, Eric W.

AU - Jeffrey, R. Brooke

AU - Yee, Judy

AU - Zalis, Michael E.

AU - Margolis, Daniel

AU - Paik, David S.

AU - Sherbondy, Anthony J.

AU - Sundaram, Padmavathi

AU - Beaulieu, Christopher F.

PY - 2006/6/1

Y1 - 2006/6/1

N2 - Purpose: To retrospectively determine if three-dimensional (3D) viewing improves radiologists' accuracy in classifying true-positive (TP) and false-positive (FP) polyp candidates identified with computer-aided detection (CAD) and to determine candidate polyp features that are associated with classification accuracy, with known polyps serving as the reference standard. Materials and Methods: Institutional review board approval and informed consent were obtained; this study was HIPAA compliant. Forty-seven computed tomographic (CT) colonography data sets were obtained in 26 men and 10 women (age range, 42-76 years). Four radiologists classified 705 polyp candidates (53 TP candidates, 652 FP candidates) identified with CAD; initially, only two-dimensional images were used, but these were later supplemented with 3D rendering. Another radiologist unblinded to colonoscopy findings characterized the features of each candidate, assessed co-Ion distention and preparation, and defined the true nature of FP candidates. Receiver operating characteristic curves were used to compare readers' performance, and repeated-measures analysis of variance was used to test features that affect interpretation. Results: Use of 3D viewing improved classification accuracy for three readers and increased the area under the receiver operating characteristic curve to 0.96-0.97 (P < .001). For TP candidates, maximum polyp width (P = .038), polyp height (P = .019), and preparation (P = .004) significantly affected accuracy. For FP candidates, colonic segment (P = .007), attenuation (P < .001), surface smoothness (P < .001), distention (P = .034), preparation (P < .001), and true nature of candidate lesions (P < .001) significantly affected accuracy. Conclusion: Use of 3D viewing increases reader accuracy in the classification of polyp candidates identified with CAD. Polyp size and examination quality are significantly associated with accuracy.

AB - Purpose: To retrospectively determine if three-dimensional (3D) viewing improves radiologists' accuracy in classifying true-positive (TP) and false-positive (FP) polyp candidates identified with computer-aided detection (CAD) and to determine candidate polyp features that are associated with classification accuracy, with known polyps serving as the reference standard. Materials and Methods: Institutional review board approval and informed consent were obtained; this study was HIPAA compliant. Forty-seven computed tomographic (CT) colonography data sets were obtained in 26 men and 10 women (age range, 42-76 years). Four radiologists classified 705 polyp candidates (53 TP candidates, 652 FP candidates) identified with CAD; initially, only two-dimensional images were used, but these were later supplemented with 3D rendering. Another radiologist unblinded to colonoscopy findings characterized the features of each candidate, assessed co-Ion distention and preparation, and defined the true nature of FP candidates. Receiver operating characteristic curves were used to compare readers' performance, and repeated-measures analysis of variance was used to test features that affect interpretation. Results: Use of 3D viewing improved classification accuracy for three readers and increased the area under the receiver operating characteristic curve to 0.96-0.97 (P < .001). For TP candidates, maximum polyp width (P = .038), polyp height (P = .019), and preparation (P = .004) significantly affected accuracy. For FP candidates, colonic segment (P = .007), attenuation (P < .001), surface smoothness (P < .001), distention (P = .034), preparation (P < .001), and true nature of candidate lesions (P < .001) significantly affected accuracy. Conclusion: Use of 3D viewing increases reader accuracy in the classification of polyp candidates identified with CAD. Polyp size and examination quality are significantly associated with accuracy.

UR - http://www.scopus.com/inward/record.url?scp=33744794891&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33744794891&partnerID=8YFLogxK

U2 - 10.1148/radiol.2393050418

DO - 10.1148/radiol.2393050418

M3 - Review article

VL - 239

SP - 768

EP - 776

JO - Radiology

JF - Radiology

SN - 0033-8419

IS - 3

ER -