Improved quantitative parameter estimation for prostate T2 relaxometry using convolutional neural networks

被引:1
作者
Bolan, Patrick J. [1 ,2 ]
Saunders, Sara L. [3 ]
Kay, Kendrick [1 ,2 ]
Gross, Mitchell [3 ]
Akcakaya, Mehmet [1 ,4 ]
Metzger, Gregory J. [1 ,2 ]
机构
[1] Univ Minnesota, Ctr Magnet Resonance Res, 2021 6th St SE, Minneapolis, MN 55455 USA
[2] Univ Minnesota, Dept Radiol, Minneapolis, MN 55455 USA
[3] Univ Minnesota, Dept Biomed Engn, Minneapolis, MN USA
[4] Univ Minnesota, Dept Elect & Comp Engn, Minneapolis, MN USA
基金
美国国家卫生研究院;
关键词
Magnetic resonance imaging; Prostate; Relaxometry; Neural networks; T-2; mapping; ARTICULAR-CARTILAGE; RECONSTRUCTION; T2; NOISE; SIMULATION; ACCURACY; IMPACT;
D O I
10.1007/s10334-024-01186-3
中图分类号
R8 [特种医学]; R445 [影像诊断学];
学科分类号
1002 ; 100207 ; 1009 ;
摘要
Objective Quantitative parameter mapping conventionally relies on curve fitting techniques to estimate parameters from magnetic resonance image series. This study compares conventional curve fitting techniques to methods using neural networks (NN) for measuring T-2 in the prostate. Materials and methods Large physics-based synthetic datasets simulating T(2 )mapping acquisitions were generated for training NNs and for quantitative performance comparisons. Four combinations of different NN architectures and training corpora were implemented and compared with four different curve fitting strategies. All methods were compared quantitatively using synthetic data with known ground truth, and further compared on in vivo test data, with and without noise augmentation, to evaluate feasibility and noise robustness. Results In the evaluation on synthetic data, a convolutional neural network (CNN), trained in a supervised fashion using synthetic data generated from naturalistic images, showed the highest overall accuracy and precision amongst the methods. On in vivo data, this best performing method produced low-noise T-2 maps and showed the least deterioration with increasing input noise levels. Discussion This study showed that a CNN, trained with synthetic data in a supervised manner, may provide superior T-2 estimation performance compared to conventional curve fitting, especially in low signal-to-noise regions.
引用
收藏
页码:721 / 735
页数:15
相关论文
共 57 条
[1]   Deep learning how to fit an intravoxel incoherent motion model to diffusion-weighted MRI [J].
Barbieri, Sebastiano ;
Gurney-Champion, Oliver J. ;
Klaassen, Remy ;
Thoeny, Harriet C. .
MAGNETIC RESONANCE IN MEDICINE, 2020, 83 (01) :312-321
[2]   Rapid and Accurate T2 Mapping from Multi-Spin-Echo Data Using Bloch-Simulation-Based Reconstruction [J].
Ben-Eliezer, Noam ;
Sodickson, Daniel K. ;
Block, Kai Tobias .
MAGNETIC RESONANCE IN MEDICINE, 2015, 73 (02) :809-817
[3]   Diffusion parameter mapping with the combined intravoxel incoherent motion and kurtosis model using artificial neural networks at 3 T [J].
Bertleff, Marco ;
Domsch, Sebastian ;
Weingaertner, Sebastian ;
Zapp, Jascha ;
O'Brien, Kieran ;
Barth, Markus ;
Schad, Lothar R. .
NMR IN BIOMEDICINE, 2017, 30 (12)
[4]   Model-Based Iterative Reconstruction for Radial Fast Spin-Echo MRI [J].
Block, Kai Tobias ;
Uecker, Martin ;
Frahm, Jens .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 2009, 28 (11) :1759-1769
[5]  
Bolan PJ, 2023, IMPROVED QUANTITATIV, DOI [10.1011/2023.01.11.23284194, DOI 10.1011/2023.01.11.23284194]
[6]   T-2 maximum likelihood estimation from multiple spin-echo magnitude images [J].
Bonny, JM ;
Zanca, M ;
Boire, JY ;
Veyre, A .
MAGNETIC RESONANCE IN MEDICINE, 1996, 36 (02) :287-293
[7]   Incorporation of Rician Noise in the Analysis of Biexponential Transverse Relaxation in Cartilage Using a Multiple Gradient Echo Sequence at 3 and 7 Tesla [J].
Bouhrara, Mustapha ;
Reiter, David A. ;
Celik, Hasan ;
Bonny, Jean-Marie ;
Lukas, Vanessa ;
Fishbein, Kenneth W. ;
Spencer, Richard G. .
MAGNETIC RESONANCE IN MEDICINE, 2015, 73 (01) :352-366
[8]   Single-shot T2 mapping using overlapping-echo detachment planar imaging and a deep convolutional neural network [J].
Cai, Congbo ;
Wang, Chao ;
Zeng, Yiqing ;
Cai, Shuhui ;
Liang, Dong ;
Wu, Yawen ;
Chen, Zhong ;
Ding, Xinghao ;
Zhong, Jianhui .
MAGNETIC RESONANCE IN MEDICINE, 2018, 80 (05) :2202-2214
[9]   MR fingerprinting Deep RecOnstruction NEtwork (DRONE) [J].
Cohen, Ouri ;
Zhu, Bo ;
Rosen, Matthew S. .
MAGNETIC RESONANCE IN MEDICINE, 2018, 80 (03) :885-894
[10]  
Deng J, 2009, PROC CVPR IEEE, P248, DOI 10.1109/CVPRW.2009.5206848