Surface curvature estimation for automatic colonic polyp detection

Adam Huang, Ronald M. Summers, Amy K. Hara

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

42 Citations (Scopus)

Abstract

Colonie polyps are growths on the inner wall of the colon. They appear like elliptical protrusions which can be detected by curvature-derived shape discriminators. For reasons of computation efficiency, much of the past work in computer-aided diagnostic CT colonography adopted kernel-based convolution methods in curvature estimation. However, kernel methods can yield erroneous results at thin structures where the gradient diminishes. In this paper, we investigate three surface patch fitting methods: Cubic B-spline, paraboloid, and quadratic polynomials. This "patch" approach is based on the fact that a surface can be re-oriented such that it can be approximated by a bivariate function locally. These patch methods are evaluated by synthesized data with various orientations and sampling sizes. We find that the cubic spline method performs best regardless of large orientation variances. Cubic spline and quadratic polynomial methods perform equally well for large samples while the latter performs better for small ones. Based on the performance evaluation, we propose a new, two-stage curvature estimation method. The cubic spline fitting is performed first for its insensitivity to orientation. If the spline fitting errs by more than a preset value (indicating high surface tortuosity), a small data sample is fitted by a quadratic function. The evaluation is performed on 29 patients (58 data sets). With 88.7% sensitivity, the average number of false positives per data set is reduced by 44.5% from 33.5 (kernel method) to 18.6 (new method).

Original languageEnglish (US)
Title of host publicationProgress in Biomedical Optics and Imaging - Proceedings of SPIE
EditorsA.A. Amini, A. Manduca
Pages393-402
Number of pages10
Volume5746
EditionI
DOIs
StatePublished - 2005
EventMedical Imaging 2005 - Physiology, Function, and Structure from Medical Images - San Diego, CA, United States
Duration: Feb 13 2005Feb 15 2005

Other

OtherMedical Imaging 2005 - Physiology, Function, and Structure from Medical Images
CountryUnited States
CitySan Diego, CA
Period2/13/052/15/05

Fingerprint

Splines
Polynomials
Discriminators
Convolution
Sampling

Keywords

  • Computer-aided diagnosis
  • CT colonography
  • Curvature estimation; surface fitting
  • Polyp detection
  • Virtual colonoscopy
  • Volume data

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Huang, A., Summers, R. M., & Hara, A. K. (2005). Surface curvature estimation for automatic colonic polyp detection. In A. A. Amini, & A. Manduca (Eds.), Progress in Biomedical Optics and Imaging - Proceedings of SPIE (I ed., Vol. 5746, pp. 393-402). [43] https://doi.org/10.1117/12.594644

Surface curvature estimation for automatic colonic polyp detection. / Huang, Adam; Summers, Ronald M.; Hara, Amy K.

Progress in Biomedical Optics and Imaging - Proceedings of SPIE. ed. / A.A. Amini; A. Manduca. Vol. 5746 I. ed. 2005. p. 393-402 43.

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

Huang, A, Summers, RM & Hara, AK 2005, Surface curvature estimation for automatic colonic polyp detection. in AA Amini & A Manduca (eds), Progress in Biomedical Optics and Imaging - Proceedings of SPIE. I edn, vol. 5746, 43, pp. 393-402, Medical Imaging 2005 - Physiology, Function, and Structure from Medical Images, San Diego, CA, United States, 2/13/05. https://doi.org/10.1117/12.594644
Huang A, Summers RM, Hara AK. Surface curvature estimation for automatic colonic polyp detection. In Amini AA, Manduca A, editors, Progress in Biomedical Optics and Imaging - Proceedings of SPIE. I ed. Vol. 5746. 2005. p. 393-402. 43 https://doi.org/10.1117/12.594644
Huang, Adam ; Summers, Ronald M. ; Hara, Amy K. / Surface curvature estimation for automatic colonic polyp detection. Progress in Biomedical Optics and Imaging - Proceedings of SPIE. editor / A.A. Amini ; A. Manduca. Vol. 5746 I. ed. 2005. pp. 393-402
@inproceedings{8c6094f5e9624ac791139896bf09d3a0,
title = "Surface curvature estimation for automatic colonic polyp detection",
abstract = "Colonie polyps are growths on the inner wall of the colon. They appear like elliptical protrusions which can be detected by curvature-derived shape discriminators. For reasons of computation efficiency, much of the past work in computer-aided diagnostic CT colonography adopted kernel-based convolution methods in curvature estimation. However, kernel methods can yield erroneous results at thin structures where the gradient diminishes. In this paper, we investigate three surface patch fitting methods: Cubic B-spline, paraboloid, and quadratic polynomials. This {"}patch{"} approach is based on the fact that a surface can be re-oriented such that it can be approximated by a bivariate function locally. These patch methods are evaluated by synthesized data with various orientations and sampling sizes. We find that the cubic spline method performs best regardless of large orientation variances. Cubic spline and quadratic polynomial methods perform equally well for large samples while the latter performs better for small ones. Based on the performance evaluation, we propose a new, two-stage curvature estimation method. The cubic spline fitting is performed first for its insensitivity to orientation. If the spline fitting errs by more than a preset value (indicating high surface tortuosity), a small data sample is fitted by a quadratic function. The evaluation is performed on 29 patients (58 data sets). With 88.7{\%} sensitivity, the average number of false positives per data set is reduced by 44.5{\%} from 33.5 (kernel method) to 18.6 (new method).",
keywords = "Computer-aided diagnosis, CT colonography, Curvature estimation; surface fitting, Polyp detection, Virtual colonoscopy, Volume data",
author = "Adam Huang and Summers, {Ronald M.} and Hara, {Amy K.}",
year = "2005",
doi = "10.1117/12.594644",
language = "English (US)",
volume = "5746",
pages = "393--402",
editor = "A.A. Amini and A. Manduca",
booktitle = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",
edition = "I",

}

TY - GEN

T1 - Surface curvature estimation for automatic colonic polyp detection

AU - Huang, Adam

AU - Summers, Ronald M.

AU - Hara, Amy K.

PY - 2005

Y1 - 2005

N2 - Colonie polyps are growths on the inner wall of the colon. They appear like elliptical protrusions which can be detected by curvature-derived shape discriminators. For reasons of computation efficiency, much of the past work in computer-aided diagnostic CT colonography adopted kernel-based convolution methods in curvature estimation. However, kernel methods can yield erroneous results at thin structures where the gradient diminishes. In this paper, we investigate three surface patch fitting methods: Cubic B-spline, paraboloid, and quadratic polynomials. This "patch" approach is based on the fact that a surface can be re-oriented such that it can be approximated by a bivariate function locally. These patch methods are evaluated by synthesized data with various orientations and sampling sizes. We find that the cubic spline method performs best regardless of large orientation variances. Cubic spline and quadratic polynomial methods perform equally well for large samples while the latter performs better for small ones. Based on the performance evaluation, we propose a new, two-stage curvature estimation method. The cubic spline fitting is performed first for its insensitivity to orientation. If the spline fitting errs by more than a preset value (indicating high surface tortuosity), a small data sample is fitted by a quadratic function. The evaluation is performed on 29 patients (58 data sets). With 88.7% sensitivity, the average number of false positives per data set is reduced by 44.5% from 33.5 (kernel method) to 18.6 (new method).

AB - Colonie polyps are growths on the inner wall of the colon. They appear like elliptical protrusions which can be detected by curvature-derived shape discriminators. For reasons of computation efficiency, much of the past work in computer-aided diagnostic CT colonography adopted kernel-based convolution methods in curvature estimation. However, kernel methods can yield erroneous results at thin structures where the gradient diminishes. In this paper, we investigate three surface patch fitting methods: Cubic B-spline, paraboloid, and quadratic polynomials. This "patch" approach is based on the fact that a surface can be re-oriented such that it can be approximated by a bivariate function locally. These patch methods are evaluated by synthesized data with various orientations and sampling sizes. We find that the cubic spline method performs best regardless of large orientation variances. Cubic spline and quadratic polynomial methods perform equally well for large samples while the latter performs better for small ones. Based on the performance evaluation, we propose a new, two-stage curvature estimation method. The cubic spline fitting is performed first for its insensitivity to orientation. If the spline fitting errs by more than a preset value (indicating high surface tortuosity), a small data sample is fitted by a quadratic function. The evaluation is performed on 29 patients (58 data sets). With 88.7% sensitivity, the average number of false positives per data set is reduced by 44.5% from 33.5 (kernel method) to 18.6 (new method).

KW - Computer-aided diagnosis

KW - CT colonography

KW - Curvature estimation; surface fitting

KW - Polyp detection

KW - Virtual colonoscopy

KW - Volume data

UR - http://www.scopus.com/inward/record.url?scp=23844453738&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=23844453738&partnerID=8YFLogxK

U2 - 10.1117/12.594644

DO - 10.1117/12.594644

M3 - Conference contribution

AN - SCOPUS:23844453738

VL - 5746

SP - 393

EP - 402

BT - Progress in Biomedical Optics and Imaging - Proceedings of SPIE

A2 - Amini, A.A.

A2 - Manduca, A.

ER -