MR image synthesis by contrast learning on neighborhood ensembles

Amod Jog, Aaron Carass, Snehashis Roy, Dzung L. Pham, Jerry Ladd Prince

Research output: Contribution to journalArticle

Abstract

Automatic processing of magnetic resonance images is a vital part of neuroscience research. Yet even the best and most widely used medical image processing methods will not produce consistent results when their input images are acquired with different pulse sequences. Although intensity standardization and image synthesis methods have been introduced to address this problem, their performance remains dependent on knowledge and consistency of the pulse sequences used to acquire the images. In this paper, an image synthesis approach that first estimates the pulse sequence parameters of the subject image is presented. The estimated parameters are then used with a collection of atlas or training images to generate a new atlas image having the same contrast as the subject image. This additional image provides an ideal source from which to synthesize any other target pulse sequence image contained in the atlas. In particular, a nonlinear regression intensity mapping is trained from the new atlas image to the target atlas image and then applied to the subject image to yield the particular target pulse sequence within the atlas. Both intensity standardization and synthesis of missing tissue contrasts can be achieved using this framework. The approach was evaluated on both simulated and real data, and shown to be superior in both intensity standardization and synthesis to other established methods.

Original languageEnglish (US)
Pages (from-to)63-76
Number of pages14
JournalMedical Image Analysis
Volume24
Issue number1
DOIs
StatePublished - Aug 1 2015

Fingerprint

Atlases
Learning
Standardization
Medical image processing
Magnetic resonance
Tissue
Neurosciences
Magnetic Resonance Spectroscopy
Processing
Research

Keywords

  • Brain
  • Magnetic resonance imaging
  • Pulse sequence
  • Synthesis

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition
  • Radiology Nuclear Medicine and imaging
  • Health Informatics
  • Radiological and Ultrasound Technology

Cite this

MR image synthesis by contrast learning on neighborhood ensembles. / Jog, Amod; Carass, Aaron; Roy, Snehashis; Pham, Dzung L.; Prince, Jerry Ladd.

In: Medical Image Analysis, Vol. 24, No. 1, 01.08.2015, p. 63-76.

Research output: Contribution to journalArticle

Jog, Amod ; Carass, Aaron ; Roy, Snehashis ; Pham, Dzung L. ; Prince, Jerry Ladd. / MR image synthesis by contrast learning on neighborhood ensembles. In: Medical Image Analysis. 2015 ; Vol. 24, No. 1. pp. 63-76.
@article{5913ddce0e714e73aad655756548e1cf,
title = "MR image synthesis by contrast learning on neighborhood ensembles",
abstract = "Automatic processing of magnetic resonance images is a vital part of neuroscience research. Yet even the best and most widely used medical image processing methods will not produce consistent results when their input images are acquired with different pulse sequences. Although intensity standardization and image synthesis methods have been introduced to address this problem, their performance remains dependent on knowledge and consistency of the pulse sequences used to acquire the images. In this paper, an image synthesis approach that first estimates the pulse sequence parameters of the subject image is presented. The estimated parameters are then used with a collection of atlas or training images to generate a new atlas image having the same contrast as the subject image. This additional image provides an ideal source from which to synthesize any other target pulse sequence image contained in the atlas. In particular, a nonlinear regression intensity mapping is trained from the new atlas image to the target atlas image and then applied to the subject image to yield the particular target pulse sequence within the atlas. Both intensity standardization and synthesis of missing tissue contrasts can be achieved using this framework. The approach was evaluated on both simulated and real data, and shown to be superior in both intensity standardization and synthesis to other established methods.",
keywords = "Brain, Magnetic resonance imaging, Pulse sequence, Synthesis",
author = "Amod Jog and Aaron Carass and Snehashis Roy and Pham, {Dzung L.} and Prince, {Jerry Ladd}",
year = "2015",
month = "8",
day = "1",
doi = "10.1016/j.media.2015.05.002",
language = "English (US)",
volume = "24",
pages = "63--76",
journal = "Medical Image Analysis",
issn = "1361-8415",
publisher = "Elsevier",
number = "1",

}

TY - JOUR

T1 - MR image synthesis by contrast learning on neighborhood ensembles

AU - Jog, Amod

AU - Carass, Aaron

AU - Roy, Snehashis

AU - Pham, Dzung L.

AU - Prince, Jerry Ladd

PY - 2015/8/1

Y1 - 2015/8/1

N2 - Automatic processing of magnetic resonance images is a vital part of neuroscience research. Yet even the best and most widely used medical image processing methods will not produce consistent results when their input images are acquired with different pulse sequences. Although intensity standardization and image synthesis methods have been introduced to address this problem, their performance remains dependent on knowledge and consistency of the pulse sequences used to acquire the images. In this paper, an image synthesis approach that first estimates the pulse sequence parameters of the subject image is presented. The estimated parameters are then used with a collection of atlas or training images to generate a new atlas image having the same contrast as the subject image. This additional image provides an ideal source from which to synthesize any other target pulse sequence image contained in the atlas. In particular, a nonlinear regression intensity mapping is trained from the new atlas image to the target atlas image and then applied to the subject image to yield the particular target pulse sequence within the atlas. Both intensity standardization and synthesis of missing tissue contrasts can be achieved using this framework. The approach was evaluated on both simulated and real data, and shown to be superior in both intensity standardization and synthesis to other established methods.

AB - Automatic processing of magnetic resonance images is a vital part of neuroscience research. Yet even the best and most widely used medical image processing methods will not produce consistent results when their input images are acquired with different pulse sequences. Although intensity standardization and image synthesis methods have been introduced to address this problem, their performance remains dependent on knowledge and consistency of the pulse sequences used to acquire the images. In this paper, an image synthesis approach that first estimates the pulse sequence parameters of the subject image is presented. The estimated parameters are then used with a collection of atlas or training images to generate a new atlas image having the same contrast as the subject image. This additional image provides an ideal source from which to synthesize any other target pulse sequence image contained in the atlas. In particular, a nonlinear regression intensity mapping is trained from the new atlas image to the target atlas image and then applied to the subject image to yield the particular target pulse sequence within the atlas. Both intensity standardization and synthesis of missing tissue contrasts can be achieved using this framework. The approach was evaluated on both simulated and real data, and shown to be superior in both intensity standardization and synthesis to other established methods.

KW - Brain

KW - Magnetic resonance imaging

KW - Pulse sequence

KW - Synthesis

UR - http://www.scopus.com/inward/record.url?scp=84930958495&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84930958495&partnerID=8YFLogxK

U2 - 10.1016/j.media.2015.05.002

DO - 10.1016/j.media.2015.05.002

M3 - Article

VL - 24

SP - 63

EP - 76

JO - Medical Image Analysis

JF - Medical Image Analysis

SN - 1361-8415

IS - 1

ER -