Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award)

Weimin Zhou; Mark A. Anastasio

doi:10.1117/12.2293772

Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award)

Weimin Zhou
, Mark A. Anastasio

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

21 Scopus citations

Abstract

It has been advocated that task-based measures of image quality (IQ) should be employed to evaluate and optimize imaging systems. Task-based measures of IQ quantify the performance of an observer on a medically relevant task. The Bayesian Ideal Observer (IO), which employs complete statistical information of the object and noise, achieves the upper limit of the performance for a binary signal classification task. However, computing the IO performance is generally analytically intractable and can be computationally burdensome when Markov-chain Monte Carlo (MCMC) techniques are employed. In this paper, supervised learning with convolutional neural networks (CNNs) is employed to approximate the IO test statistics for a signal-known-exactly and background-known-exactly (SKE/BKE) binary detection task. The receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) are compared to those produced by the analytically computed IO. The advantages of the proposed supervised learning approach for approximating the IO are demonstrated.

Original language	English (US)
Title of host publication	Medical Imaging 2018
Subtitle of host publication	Image Perception, Observer Performance, and Technology Assessment
Editors	Frank W. Samuelson, Robert M. Nishikawa
Publisher	SPIE
ISBN (Electronic)	9781510616431
DOIs	https://doi.org/10.1117/12.2293772
State	Published - 2018
Externally published	Yes
Event	Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment - Houston, United States Duration: Feb 11 2018 → Feb 12 2018

Publication series

Name	Progress in Biomedical Optics and Imaging - Proceedings of SPIE
Volume	10577
ISSN (Print)	1605-7422

Conference

Conference	Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment
Country/Territory	United States
City	Houston
Period	2/11/18 → 2/12/18

Keywords

Bayesian Ideal Observer
convolutional neural networks
signal detection theory
Supervised learning

ASJC Scopus subject areas

Electronic, Optical and Magnetic Materials
Atomic and Molecular Physics, and Optics
Biomaterials
Radiology Nuclear Medicine and imaging

Access to Document

10.1117/12.2293772

Cite this

Zhou, W., & Anastasio, M. A. (2018). Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award). In F. W. Samuelson, & R. M. Nishikawa (Eds.), Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment Article 1057719 (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 10577). SPIE. https://doi.org/10.1117/12.2293772

Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award). / Zhou, Weimin; Anastasio, Mark A.
Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment. ed. / Frank W. Samuelson; Robert M. Nishikawa. SPIE, 2018. 1057719 (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 10577).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Zhou, W & Anastasio, MA 2018, Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award). in FW Samuelson & RM Nishikawa (eds), Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment., 1057719, Progress in Biomedical Optics and Imaging - Proceedings of SPIE, vol. 10577, SPIE, Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment, Houston, United States, 2/11/18. https://doi.org/10.1117/12.2293772

Zhou W, Anastasio MA. Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award). In Samuelson FW, Nishikawa RM, editors, Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment. SPIE. 2018. 1057719. (Progress in Biomedical Optics and Imaging - Proceedings of SPIE). doi: 10.1117/12.2293772

Zhou, Weimin ; Anastasio, Mark A. / Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award). Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment. editor / Frank W. Samuelson ; Robert M. Nishikawa. SPIE, 2018. (Progress in Biomedical Optics and Imaging - Proceedings of SPIE).

@inproceedings{dcf1d7da5ce14078bf650438bd0ac459,

title = "Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award)",

abstract = "It has been advocated that task-based measures of image quality (IQ) should be employed to evaluate and optimize imaging systems. Task-based measures of IQ quantify the performance of an observer on a medically relevant task. The Bayesian Ideal Observer (IO), which employs complete statistical information of the object and noise, achieves the upper limit of the performance for a binary signal classification task. However, computing the IO performance is generally analytically intractable and can be computationally burdensome when Markov-chain Monte Carlo (MCMC) techniques are employed. In this paper, supervised learning with convolutional neural networks (CNNs) is employed to approximate the IO test statistics for a signal-known-exactly and background-known-exactly (SKE/BKE) binary detection task. The receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) are compared to those produced by the analytically computed IO. The advantages of the proposed supervised learning approach for approximating the IO are demonstrated.",

keywords = "Bayesian Ideal Observer, convolutional neural networks, signal detection theory, Supervised learning",

author = "Weimin Zhou and Anastasio, \{Mark A.\}",

note = "Publisher Copyright: {\textcopyright} 2018 SPIE.; Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment ; Conference date: 11-02-2018 Through 12-02-2018",

year = "2018",

doi = "10.1117/12.2293772",

language = "English (US)",

series = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",

publisher = "SPIE",

editor = "Samuelson, \{Frank W.\} and Nishikawa, \{Robert M.\}",

booktitle = "Medical Imaging 2018",

}

TY - GEN

T1 - Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award)

AU - Zhou, Weimin

AU - Anastasio, Mark A.

PY - 2018

Y1 - 2018

N2 - It has been advocated that task-based measures of image quality (IQ) should be employed to evaluate and optimize imaging systems. Task-based measures of IQ quantify the performance of an observer on a medically relevant task. The Bayesian Ideal Observer (IO), which employs complete statistical information of the object and noise, achieves the upper limit of the performance for a binary signal classification task. However, computing the IO performance is generally analytically intractable and can be computationally burdensome when Markov-chain Monte Carlo (MCMC) techniques are employed. In this paper, supervised learning with convolutional neural networks (CNNs) is employed to approximate the IO test statistics for a signal-known-exactly and background-known-exactly (SKE/BKE) binary detection task. The receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) are compared to those produced by the analytically computed IO. The advantages of the proposed supervised learning approach for approximating the IO are demonstrated.

AB - It has been advocated that task-based measures of image quality (IQ) should be employed to evaluate and optimize imaging systems. Task-based measures of IQ quantify the performance of an observer on a medically relevant task. The Bayesian Ideal Observer (IO), which employs complete statistical information of the object and noise, achieves the upper limit of the performance for a binary signal classification task. However, computing the IO performance is generally analytically intractable and can be computationally burdensome when Markov-chain Monte Carlo (MCMC) techniques are employed. In this paper, supervised learning with convolutional neural networks (CNNs) is employed to approximate the IO test statistics for a signal-known-exactly and background-known-exactly (SKE/BKE) binary detection task. The receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) are compared to those produced by the analytically computed IO. The advantages of the proposed supervised learning approach for approximating the IO are demonstrated.

KW - Bayesian Ideal Observer

KW - convolutional neural networks

KW - signal detection theory

KW - Supervised learning

UR - https://www.scopus.com/pages/publications/85047847289

UR - https://www.scopus.com/pages/publications/85047847289#tab=citedBy

U2 - 10.1117/12.2293772

DO - 10.1117/12.2293772

M3 - Conference contribution

AN - SCOPUS:85047847289

T3 - Progress in Biomedical Optics and Imaging - Proceedings of SPIE

BT - Medical Imaging 2018

A2 - Samuelson, Frank W.

A2 - Nishikawa, Robert M.

PB - SPIE

T2 - Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment

Y2 - 11 February 2018 through 12 February 2018

ER -

Learning the ideal observer for SKE detection tasks by use of convolutional neural networks (Cum Laude Poster Award)

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this