Improving self super resolution in magnetic resonance images

Sachin Goyal, Can Zhao, Amod Jog, Jerry Ladd Prince, Aaron Carass

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

Abstract

Magnetic resonance (MR) images (MRI) are routinely acquired with high in-plane resolution and lower through-plane resolution. Improving the resolution of such data can be achieved through post-processing techniques knows as super-resolution (SR), with various frameworks in existence. Many of these approaches rely on external databases from which SR methods infer relationships between low and high resolution data. The concept of self super-resolution (SSR) has been previously reported, wherein there is no external training data with the method only relying on the acquired image. The approach involves extracting image patches from the acquired image constructing new images based on regression and combining the new images by Fourier Burst Accumulation. In this work, we present four improvements to our previously reported SSR approach. We demonstrate these improvements have a significant effect on improving image quality and the measured resolution.

Original languageEnglish (US)
Title of host publicationMedical Imaging 2018
Subtitle of host publicationBiomedical Applications in Molecular, Structural, and Functional Imaging
EditorsBarjor Gimi, Andrzej Krol
PublisherSPIE
Volume10578
ISBN (Electronic)9781510616455
DOIs
StatePublished - Jan 1 2018
EventMedical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging - Houston, United States
Duration: Feb 11 2018Feb 13 2018

Other

OtherMedical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging
CountryUnited States
CityHouston
Period2/11/182/13/18

Fingerprint

Magnetic resonance
Image quality
magnetic resonance
Magnetic Resonance Spectroscopy
Post and Core Technique
Processing
Self Concept
Databases
regression analysis
bursts
education
high resolution

Keywords

  • MRI
  • super-resolution
  • unsupervised methods

ASJC Scopus subject areas

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

Cite this

Goyal, S., Zhao, C., Jog, A., Prince, J. L., & Carass, A. (2018). Improving self super resolution in magnetic resonance images. In B. Gimi, & A. Krol (Eds.), Medical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging (Vol. 10578). [1057814] SPIE. https://doi.org/10.1117/12.2295366

Improving self super resolution in magnetic resonance images. / Goyal, Sachin; Zhao, Can; Jog, Amod; Prince, Jerry Ladd; Carass, Aaron.

Medical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging. ed. / Barjor Gimi; Andrzej Krol. Vol. 10578 SPIE, 2018. 1057814.

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

Goyal, S, Zhao, C, Jog, A, Prince, JL & Carass, A 2018, Improving self super resolution in magnetic resonance images. in B Gimi & A Krol (eds), Medical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging. vol. 10578, 1057814, SPIE, Medical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging, Houston, United States, 2/11/18. https://doi.org/10.1117/12.2295366
Goyal S, Zhao C, Jog A, Prince JL, Carass A. Improving self super resolution in magnetic resonance images. In Gimi B, Krol A, editors, Medical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging. Vol. 10578. SPIE. 2018. 1057814 https://doi.org/10.1117/12.2295366
Goyal, Sachin ; Zhao, Can ; Jog, Amod ; Prince, Jerry Ladd ; Carass, Aaron. / Improving self super resolution in magnetic resonance images. Medical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging. editor / Barjor Gimi ; Andrzej Krol. Vol. 10578 SPIE, 2018.
@inproceedings{44f33540e45744668d06c3cb0a3d6303,
title = "Improving self super resolution in magnetic resonance images",
abstract = "Magnetic resonance (MR) images (MRI) are routinely acquired with high in-plane resolution and lower through-plane resolution. Improving the resolution of such data can be achieved through post-processing techniques knows as super-resolution (SR), with various frameworks in existence. Many of these approaches rely on external databases from which SR methods infer relationships between low and high resolution data. The concept of self super-resolution (SSR) has been previously reported, wherein there is no external training data with the method only relying on the acquired image. The approach involves extracting image patches from the acquired image constructing new images based on regression and combining the new images by Fourier Burst Accumulation. In this work, we present four improvements to our previously reported SSR approach. We demonstrate these improvements have a significant effect on improving image quality and the measured resolution.",
keywords = "MRI, super-resolution, unsupervised methods",
author = "Sachin Goyal and Can Zhao and Amod Jog and Prince, {Jerry Ladd} and Aaron Carass",
year = "2018",
month = "1",
day = "1",
doi = "10.1117/12.2295366",
language = "English (US)",
volume = "10578",
editor = "Barjor Gimi and Andrzej Krol",
booktitle = "Medical Imaging 2018",
publisher = "SPIE",

}

TY - GEN

T1 - Improving self super resolution in magnetic resonance images

AU - Goyal, Sachin

AU - Zhao, Can

AU - Jog, Amod

AU - Prince, Jerry Ladd

AU - Carass, Aaron

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Magnetic resonance (MR) images (MRI) are routinely acquired with high in-plane resolution and lower through-plane resolution. Improving the resolution of such data can be achieved through post-processing techniques knows as super-resolution (SR), with various frameworks in existence. Many of these approaches rely on external databases from which SR methods infer relationships between low and high resolution data. The concept of self super-resolution (SSR) has been previously reported, wherein there is no external training data with the method only relying on the acquired image. The approach involves extracting image patches from the acquired image constructing new images based on regression and combining the new images by Fourier Burst Accumulation. In this work, we present four improvements to our previously reported SSR approach. We demonstrate these improvements have a significant effect on improving image quality and the measured resolution.

AB - Magnetic resonance (MR) images (MRI) are routinely acquired with high in-plane resolution and lower through-plane resolution. Improving the resolution of such data can be achieved through post-processing techniques knows as super-resolution (SR), with various frameworks in existence. Many of these approaches rely on external databases from which SR methods infer relationships between low and high resolution data. The concept of self super-resolution (SSR) has been previously reported, wherein there is no external training data with the method only relying on the acquired image. The approach involves extracting image patches from the acquired image constructing new images based on regression and combining the new images by Fourier Burst Accumulation. In this work, we present four improvements to our previously reported SSR approach. We demonstrate these improvements have a significant effect on improving image quality and the measured resolution.

KW - MRI

KW - super-resolution

KW - unsupervised methods

UR - http://www.scopus.com/inward/record.url?scp=85049578097&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049578097&partnerID=8YFLogxK

U2 - 10.1117/12.2295366

DO - 10.1117/12.2295366

M3 - Conference contribution

AN - SCOPUS:85049578097

VL - 10578

BT - Medical Imaging 2018

A2 - Gimi, Barjor

A2 - Krol, Andrzej

PB - SPIE

ER -