SuperMAP: Superfast MR Mapping with Joint Under-sampling using Deep Combined Network

Hongyu Li; Mingrui Yang; Jeehum Kim; Chaoyi Zhang; Ruiying Liu; Peizhau Huang; Sunil Gaire; Dong Liang; Xiaoliang Zhang; Xiaojuan Li; Leslie Ying

SuperMAP: Superfast MR Mapping with Joint Under-sampling using Deep Combined Network

Hongyu Li, Mingrui Yang, Jeehum Kim, Chaoyi Zhang, Ruiying Liu, Peizhau Huang, Sunil Gaire, Dong Liang, Xiaoliang Zhang, Xiaojuan Li, Leslie Ying

Electrical and Computer Engineering

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

Abstract

This abstract presents a combined deep learning framework SuperMAP to generate MR parameter maps from very few subsampled echo images. The method combines deep residual convolutional neural networks (DRCNN) and fully connected networks (FC) to exploit the nonlinear relationship between and within the combined subsampled T1rho/T2 weighted images and the combined T1rho/T2 maps. Experimental results show that the proposed combined network is superior to single CNN network and can generate accurate T1rho and T2 maps simultaneously from only three subsampled echoes within one scan with results comparable to reference from fully sampled 8-echo images each for two separate scans.

Original language	English
Title of host publication	Unknown book
State	Published - 2021

Cite this

@inproceedings{8a455b0772544d4eab9329e7bc545057,

title = "SuperMAP: Superfast MR Mapping with Joint Under-sampling using Deep Combined Network",

abstract = "This abstract presents a combined deep learning framework SuperMAP to generate MR parameter maps from very few subsampled echo images. The method combines deep residual convolutional neural networks (DRCNN) and fully connected networks (FC) to exploit the nonlinear relationship between and within the combined subsampled T1rho/T2 weighted images and the combined T1rho/T2 maps. Experimental results show that the proposed combined network is superior to single CNN network and can generate accurate T1rho and T2 maps simultaneously from only three subsampled echoes within one scan with results comparable to reference from fully sampled 8-echo images each for two separate scans.",

author = "Hongyu Li and Mingrui Yang and Jeehum Kim and Chaoyi Zhang and Ruiying Liu and Peizhau Huang and Sunil Gaire and Dong Liang and Xiaoliang Zhang and Xiaojuan Li and Leslie Ying",

year = "2021",

language = "English",

booktitle = "Unknown book",

}

TY - GEN

T1 - SuperMAP: Superfast MR Mapping with Joint Under-sampling using Deep Combined Network

AU - Li, Hongyu

AU - Yang, Mingrui

AU - Kim, Jeehum

AU - Zhang, Chaoyi

AU - Liu, Ruiying

AU - Huang, Peizhau

AU - Gaire, Sunil

AU - Liang, Dong

AU - Zhang, Xiaoliang

AU - Li, Xiaojuan

AU - Ying, Leslie

PY - 2021

Y1 - 2021

N2 - This abstract presents a combined deep learning framework SuperMAP to generate MR parameter maps from very few subsampled echo images. The method combines deep residual convolutional neural networks (DRCNN) and fully connected networks (FC) to exploit the nonlinear relationship between and within the combined subsampled T1rho/T2 weighted images and the combined T1rho/T2 maps. Experimental results show that the proposed combined network is superior to single CNN network and can generate accurate T1rho and T2 maps simultaneously from only three subsampled echoes within one scan with results comparable to reference from fully sampled 8-echo images each for two separate scans.

AB - This abstract presents a combined deep learning framework SuperMAP to generate MR parameter maps from very few subsampled echo images. The method combines deep residual convolutional neural networks (DRCNN) and fully connected networks (FC) to exploit the nonlinear relationship between and within the combined subsampled T1rho/T2 weighted images and the combined T1rho/T2 maps. Experimental results show that the proposed combined network is superior to single CNN network and can generate accurate T1rho and T2 maps simultaneously from only three subsampled echoes within one scan with results comparable to reference from fully sampled 8-echo images each for two separate scans.

M3 - Conference contribution

BT - Unknown book

ER -

SuperMAP: Superfast MR Mapping with Joint Under-sampling using Deep Combined Network

Abstract

Fingerprint

Cite this