Automated low-contrast pattern recognition algorithm for magnetic resonance image quality assessment

Morgan O. Ehman, Zhonghao Bao, Scott O. Stiving, Mallik Kasam, Dianna Lanners, Teresa Peterson, Renee Jonsgaard, Rickey E. Carter, Kiaran Patrick McGee

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Purpose: Low contrast (LC) detectability is a common test criterion for diagnostic radiologic quality control (QC) programs. Automation of this test is desirable in order to reduce human variability and to speed up analysis. However, automation is challenging due to the complexity of the human visual perception system and the ability to create algorithms that mimic this response. This paper describes the development and testing of an automated LC detection algorithm for use in the analysis of magnetic resonance (MR) images of the American College of Radiology (ACR) QC phantom. Methods: The detection algorithm includes fuzzy logic decision processes and various edge detection methods to quantify LC detectability. Algorithm performance was first evaluated using a single LC phantom MR image with the addition of incremental zero mean Gaussian noise resulting in a total of 200 images. A c-statistic was calculated to determine the role of CNR to indicate when the algorithm would detect ten spokes. To evaluate inter-rater agreement between experienced observers and the algorithm, a blinded observer study was performed on 196 LC phantom images acquired from nine clinical MR scanners. The nine scanners included two MR manufacturers and two field strengths (1.5 T, 3.0 T). Inter-rater and algorithm-rater agreement was quantified using Krippendorff's alpha. Results: For the Gaussian noise added data, CNR ranged from 0.519 to 11.7 with CNR being considered an excellent discriminator of algorithm performance (c-statistic = 0.9777). Reviewer scoring of the clinical phantom data resulted in an inter-rater agreement of 0.673 with the agreement between observers and algorithm equal to 0.652, both of which indicate significant agreement. Conclusions: This study demonstrates that the detection of LC test patterns for MR imaging QC programs can be successfully developed and that their response can model the human visual detection system of expert MR QC readers.

Original languageEnglish (US)
JournalMedical Physics
DOIs
StateAccepted/In press - 2017

Fingerprint

Magnetic Resonance Spectroscopy
Quality Control
Automation
Fuzzy Logic
Recognition (Psychology)
Expert Systems
Visual Perception
Aptitude
Routine Diagnostic Tests
Radiology
Magnetic Resonance Imaging

Keywords

  • American College of Radiology
  • Automated detection
  • Fuzzy logic
  • Low contrast resolution
  • Magnetic resonance imaging
  • Quality control

ASJC Scopus subject areas

  • Biophysics
  • Radiology Nuclear Medicine and imaging

Cite this

Automated low-contrast pattern recognition algorithm for magnetic resonance image quality assessment. / Ehman, Morgan O.; Bao, Zhonghao; Stiving, Scott O.; Kasam, Mallik; Lanners, Dianna; Peterson, Teresa; Jonsgaard, Renee; Carter, Rickey E.; McGee, Kiaran Patrick.

In: Medical Physics, 2017.

Research output: Contribution to journalArticle

Ehman, Morgan O. ; Bao, Zhonghao ; Stiving, Scott O. ; Kasam, Mallik ; Lanners, Dianna ; Peterson, Teresa ; Jonsgaard, Renee ; Carter, Rickey E. ; McGee, Kiaran Patrick. / Automated low-contrast pattern recognition algorithm for magnetic resonance image quality assessment. In: Medical Physics. 2017.
@article{e7d60f3fe0c3441d8093004e45d1f6fb,
title = "Automated low-contrast pattern recognition algorithm for magnetic resonance image quality assessment",
abstract = "Purpose: Low contrast (LC) detectability is a common test criterion for diagnostic radiologic quality control (QC) programs. Automation of this test is desirable in order to reduce human variability and to speed up analysis. However, automation is challenging due to the complexity of the human visual perception system and the ability to create algorithms that mimic this response. This paper describes the development and testing of an automated LC detection algorithm for use in the analysis of magnetic resonance (MR) images of the American College of Radiology (ACR) QC phantom. Methods: The detection algorithm includes fuzzy logic decision processes and various edge detection methods to quantify LC detectability. Algorithm performance was first evaluated using a single LC phantom MR image with the addition of incremental zero mean Gaussian noise resulting in a total of 200 images. A c-statistic was calculated to determine the role of CNR to indicate when the algorithm would detect ten spokes. To evaluate inter-rater agreement between experienced observers and the algorithm, a blinded observer study was performed on 196 LC phantom images acquired from nine clinical MR scanners. The nine scanners included two MR manufacturers and two field strengths (1.5 T, 3.0 T). Inter-rater and algorithm-rater agreement was quantified using Krippendorff's alpha. Results: For the Gaussian noise added data, CNR ranged from 0.519 to 11.7 with CNR being considered an excellent discriminator of algorithm performance (c-statistic = 0.9777). Reviewer scoring of the clinical phantom data resulted in an inter-rater agreement of 0.673 with the agreement between observers and algorithm equal to 0.652, both of which indicate significant agreement. Conclusions: This study demonstrates that the detection of LC test patterns for MR imaging QC programs can be successfully developed and that their response can model the human visual detection system of expert MR QC readers.",
keywords = "American College of Radiology, Automated detection, Fuzzy logic, Low contrast resolution, Magnetic resonance imaging, Quality control",
author = "Ehman, {Morgan O.} and Zhonghao Bao and Stiving, {Scott O.} and Mallik Kasam and Dianna Lanners and Teresa Peterson and Renee Jonsgaard and Carter, {Rickey E.} and McGee, {Kiaran Patrick}",
year = "2017",
doi = "10.1002/mp.12370",
language = "English (US)",
journal = "Medical Physics",
issn = "0094-2405",
publisher = "AAPM - American Association of Physicists in Medicine",

}

TY - JOUR

T1 - Automated low-contrast pattern recognition algorithm for magnetic resonance image quality assessment

AU - Ehman, Morgan O.

AU - Bao, Zhonghao

AU - Stiving, Scott O.

AU - Kasam, Mallik

AU - Lanners, Dianna

AU - Peterson, Teresa

AU - Jonsgaard, Renee

AU - Carter, Rickey E.

AU - McGee, Kiaran Patrick

PY - 2017

Y1 - 2017

N2 - Purpose: Low contrast (LC) detectability is a common test criterion for diagnostic radiologic quality control (QC) programs. Automation of this test is desirable in order to reduce human variability and to speed up analysis. However, automation is challenging due to the complexity of the human visual perception system and the ability to create algorithms that mimic this response. This paper describes the development and testing of an automated LC detection algorithm for use in the analysis of magnetic resonance (MR) images of the American College of Radiology (ACR) QC phantom. Methods: The detection algorithm includes fuzzy logic decision processes and various edge detection methods to quantify LC detectability. Algorithm performance was first evaluated using a single LC phantom MR image with the addition of incremental zero mean Gaussian noise resulting in a total of 200 images. A c-statistic was calculated to determine the role of CNR to indicate when the algorithm would detect ten spokes. To evaluate inter-rater agreement between experienced observers and the algorithm, a blinded observer study was performed on 196 LC phantom images acquired from nine clinical MR scanners. The nine scanners included two MR manufacturers and two field strengths (1.5 T, 3.0 T). Inter-rater and algorithm-rater agreement was quantified using Krippendorff's alpha. Results: For the Gaussian noise added data, CNR ranged from 0.519 to 11.7 with CNR being considered an excellent discriminator of algorithm performance (c-statistic = 0.9777). Reviewer scoring of the clinical phantom data resulted in an inter-rater agreement of 0.673 with the agreement between observers and algorithm equal to 0.652, both of which indicate significant agreement. Conclusions: This study demonstrates that the detection of LC test patterns for MR imaging QC programs can be successfully developed and that their response can model the human visual detection system of expert MR QC readers.

AB - Purpose: Low contrast (LC) detectability is a common test criterion for diagnostic radiologic quality control (QC) programs. Automation of this test is desirable in order to reduce human variability and to speed up analysis. However, automation is challenging due to the complexity of the human visual perception system and the ability to create algorithms that mimic this response. This paper describes the development and testing of an automated LC detection algorithm for use in the analysis of magnetic resonance (MR) images of the American College of Radiology (ACR) QC phantom. Methods: The detection algorithm includes fuzzy logic decision processes and various edge detection methods to quantify LC detectability. Algorithm performance was first evaluated using a single LC phantom MR image with the addition of incremental zero mean Gaussian noise resulting in a total of 200 images. A c-statistic was calculated to determine the role of CNR to indicate when the algorithm would detect ten spokes. To evaluate inter-rater agreement between experienced observers and the algorithm, a blinded observer study was performed on 196 LC phantom images acquired from nine clinical MR scanners. The nine scanners included two MR manufacturers and two field strengths (1.5 T, 3.0 T). Inter-rater and algorithm-rater agreement was quantified using Krippendorff's alpha. Results: For the Gaussian noise added data, CNR ranged from 0.519 to 11.7 with CNR being considered an excellent discriminator of algorithm performance (c-statistic = 0.9777). Reviewer scoring of the clinical phantom data resulted in an inter-rater agreement of 0.673 with the agreement between observers and algorithm equal to 0.652, both of which indicate significant agreement. Conclusions: This study demonstrates that the detection of LC test patterns for MR imaging QC programs can be successfully developed and that their response can model the human visual detection system of expert MR QC readers.

KW - American College of Radiology

KW - Automated detection

KW - Fuzzy logic

KW - Low contrast resolution

KW - Magnetic resonance imaging

KW - Quality control

UR - http://www.scopus.com/inward/record.url?scp=85021698636&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021698636&partnerID=8YFLogxK

U2 - 10.1002/mp.12370

DO - 10.1002/mp.12370

M3 - Article

C2 - 28543961

AN - SCOPUS:85021698636

JO - Medical Physics

JF - Medical Physics

SN - 0094-2405

ER -