Deep learning how to fit an intravoxel incoherent motion model to diffusion-weighted MRI

被引:106
作者
Barbieri, Sebastiano [1 ]
Gurney-Champion, Oliver J. [2 ,3 ]
Klaassen, Remy [4 ,5 ]
Thoeny, Harriet C. [6 ]
机构
[1] UNSW, Ctr Big Data Res Hlth, Sydney, NSW 2052, Australia
[2] Inst Canc Res, Joint Dept Phys, London, England
[3] Royal Marsden NHS Fdn Trust, London, England
[4] Acad Med Ctr, Canc Ctr Amsterdam, Dept Med Oncol, Amsterdam, Netherlands
[5] Acad Med Ctr, LEXOR Lab Expt Oncol & Radiobiol, Amsterdam, Netherlands
[6] HFR Fribourg Hop Cantonal, Dept Radiol, Fribourg, Switzerland
基金
瑞士国家科学基金会;
关键词
cancer; deep learning; diffusion-weighted magnetic resonance imaging; intravoxel incoherent motion; IVIM; neural network; GADOLINIUM; PERFUSION; IMPACT;
D O I
10.1002/mrm.27910
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
Purpose This prospective clinical study assesses the feasibility of training a deep neural network (DNN) for intravoxel incoherent motion (IVIM) model fitting to diffusion-weighted MRI (DW-MRI) data and evaluates its performance. Methods In May 2011, 10 male volunteers (age range, 29-53 years; mean, 37) underwent DW-MRI of the upper abdomen on 1.5T and 3.0T MR scanners. Regions of interest in the left and right liver lobe, pancreas, spleen, renal cortex, and renal medulla were delineated independently by 2 readers. DNNs were trained for IVIM model fitting using these data; results were compared to least-squares and Bayesian approaches to IVIM fitting. Intraclass correlation coefficients (ICCs) were used to assess consistency of measurements between readers. Intersubject variability was evaluated using coefficients of variation (CVs). The fitting error was calculated based on simulated data, and the average fitting time of each method was recorded. Results DNNs were trained successfully for IVIM parameter estimation. This approach was associated with high consistency between the 2 readers (ICCs between 50% and 97%), low intersubject variability of estimated parameter values (CVs between 9.2 and 28.4), and the lowest error when compared with least-squares and Bayesian approaches. Fitting by DNNs was several orders of magnitude quicker than the other methods, but the networks may need to be retrained for different acquisition protocols or imaged anatomical regions. Conclusion DNNs are recommended for accurate and robust IVIM model fitting to DW-MRI data. Suitable software is available for download.
引用
收藏
页码:312 / 321
页数:10
相关论文
共 50 条
[41]   Intravoxel incoherent motion diffusion-weighted imaging for evaluating the pancreatic perfusion in cirrhotic patients [J].
Hu, Ran ;
Zeng, Guo-Fei ;
Fang, Yu ;
Nie, Lisha ;
Liang, Hui-Lou ;
Wang, Zhi-Gang ;
Yang, Hua .
ABDOMINAL RADIOLOGY, 2024, 49 (02) :492-500
[42]   Evaluation of Intravoxel Incoherent Motion Diffusion-Weighted Magnetic Resonance Imaging for Detection of Bowel Inflammation in Patients With Crohn Disease [J].
Yang, Hai-Jing ;
Xi, Yu-Ling ;
Guan, Xue-Ni ;
Xie, Qian ;
Rong, Lan ;
Liang, Zong-Hui .
JOURNAL OF COMPUTER ASSISTED TOMOGRAPHY, 2019, 43 (05) :755-761
[43]   Intravoxel incoherent motion diffusion-weighted MRI in patients with breast cancer: Correlation with tumor stroma characteristics [J].
Yuan, Li ;
Wang, Zhengping ;
Chen, Feng ;
Qin, Xin ;
Li, Changqing ;
Zhao, Yingman ;
Yan, Chenggong ;
Wu, Yuankui ;
Hao, Peng ;
Xu, Yikai .
EUROPEAN JOURNAL OF RADIOLOGY, 2019, 120
[44]   Standard diffusion-weighted, diffusion kurtosis and intravoxel incoherent motion in differentiating invasive placentas [J].
Tao Lu ;
Mou Li ;
Yishuang Wang ;
Hang Li ;
Mingpeng Wu ;
Guotai Wang .
Archives of Gynecology and Obstetrics, 2024, 309 :503-514
[45]   Utility of intravoxel incoherent motion diffusion-weighted imaging in differentiating renal cell carcinoma from angiomyolipoma [J].
Li, Congrui ;
Liu, Yan ;
Yu, Xiaoping ;
Xie, Yu ;
Sun, Yan .
INTERNATIONAL JOURNAL OF CLINICAL AND EXPERIMENTAL MEDICINE, 2017, 10 (06) :9306-9313
[46]   Assessment of renal function using intravoxel incoherent motion diffusion-weighted imaging and dynamic contrast-enhanced MRI [J].
Bane, Octavia ;
Wagner, Mathilde ;
Zhang, Jeff L. ;
Dyvorne, Hadrien A. ;
Orton, Matthew ;
Rusinek, Henry ;
Taouli, Bachir .
JOURNAL OF MAGNETIC RESONANCE IMAGING, 2016, 44 (02) :317-326
[47]   Intravoxel Incoherent Motion Diffusion-Weighted MRI at 3.0 T Differentiates Malignant Breast Lesions From Benign Lesions and Breast Parenchyma [J].
Bokacheva, Louisa ;
Kaplan, Jennifer B. ;
Giri, Dilip D. ;
Patil, Sujata ;
Gnanasigamani, Merlin ;
Nyman, C. Gregory ;
Deasy, Joseph O. ;
Morris, Elizabeth A. ;
Thakur, Sunitha B. .
JOURNAL OF MAGNETIC RESONANCE IMAGING, 2014, 40 (04) :813-823
[48]   Standard diffusion-weighted, diffusion kurtosis and intravoxel incoherent motion in differentiating invasive placentas [J].
Lu, Tao ;
Li, Mou ;
Wang, Yishuang ;
Li, Hang ;
Wu, Mingpeng ;
Wang, Guotai .
ARCHIVES OF GYNECOLOGY AND OBSTETRICS, 2024, 309 (02) :503-514
[49]   Diffusion-Weighted Intravoxel Incoherent Motion Imaging of Renal Tumors With Histopathologic Correlation [J].
Chandarana, Hersh ;
Kang, Stella K. ;
Wong, Samson ;
Rusinek, Henry ;
Zhang, Jeff L. ;
Arizono, Shigeki ;
Huang, William C. ;
Melamed, Jonathan ;
Babb, James S. ;
Suan, Edgar F. ;
Lee, Vivian S. ;
Sigmund, Eric E. .
INVESTIGATIVE RADIOLOGY, 2012, 47 (12) :688-696
[50]   Intravoxel incoherent motion diffusion-weighted imaging in the characterization of Alzheimer’s disease [J].
Nengzhi Xia ;
Yanxuan Li ;
Yingnan Xue ;
Weikang Li ;
Zhenhua Zhang ;
Caiyun Wen ;
Jiance Li ;
Qiong Ye .
Brain Imaging and Behavior, 2022, 16 :617-626