Ensemble of expert deep neural networks for spatio-temporal denoising of contrast-enhanced MRI sequences

被引:59
作者
Benou, A. [1 ,2 ]
Veksler, R. [2 ,3 ]
Friedman, A. [2 ,3 ,4 ,5 ]
Raviv, T. Riklin [1 ,2 ]
机构
[1] Ben Gurion Univ Negev, Dept Elect & Comp Engn, Beer Sheva, Israel
[2] Ben Gurion Univ Negev, Zlotowski Ctr Neurosci, Beer Sheva, Israel
[3] Ben Gurion Univ Negev, Fac Hlth Sci, Dept Physiol & Cell Biol, Beer Sheva, Israel
[4] Dalhousie Univ, Dept Med Neurosci, Halifax, NS, Canada
[5] Dalhousie Univ, Brain Repair Ctr, Halifax, NS, Canada
基金
以色列科学基金会;
关键词
Spatio-temporal denoising; Deep neural networks (DNNs); Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI); Stacked restricted Boltzmann machine (SRBM); Blood-Brain Barrier (BBB); Pharmacokinetic parameter estimation; BLOOD-BRAIN-BARRIER; KINETIC-PARAMETERS; TRACER KINETICS; IMAGE; REPRESENTATIONS; QUANTIFICATION; ALGORITHM; LEAKAGE; MODELS; ROBUST;
D O I
10.1016/j.media.2017.07.006
中图分类号
TP18 [人工智能理论];
学科分类号
140502 [人工智能];
摘要
Dynamic contrast-enhanced MRI (DCE-MRI) is an imaging protocol where MRI scans are acquired repetitively throughout the injection of a contrast agent. The analysis of dynamic scans is widely used for the detection and quantification of blood-brain barrier (BBB) permeability. Extraction of the pharmacokinetic (PK) parameters from the DCE-MRI concentration curves allows quantitative assessment of the integrity of the BBB functionality. However, curve fitting required for the analysis of DCE-MRI data is error-prone as the dynamic scans are subject to non-white, spatially-dependent and anisotropic noise. We present a novel spatio-temporal framework based on Deep Neural Networks (DNNs) to address the DCE-MRI de noising challenges. This is accomplished by an ensemble of expert DNNs constructed as deep autoencoders, where each is trained on a specific subset of the input space to accommodate different noise characteristics and curve prototypes. Spatial dependencies of the PK dynamics are captured by incorporating the curves of neighboring voxels in the entire process. The most likely reconstructed curves are then chosen using a classifier DNN followed by a quadratic programming optimization. As clean signals (ground-truth) for training are not available, a fully automatic model for generating realistic training sets with complex nonlinear dynamics is introduced. The proposed approach has been successfully applied to full and even temporally down-sampled DCE-MRI sequences, from two different databases, of stroke and brain tumor patients, and is shown to favorably compare to state-of-the-art denoising methods. (C) 2017 Elsevier B.V. All rights reserved.
引用
收藏
页码:145 / 159
页数:15
相关论文
共 58 条
[1]
Overview and introduction: The blood-brain barrier in health and disease [J].
Abbott, N. Joan ;
Friedman, Alon .
EPILEPSIA, 2012, 53 :1-6
[2]
[Anonymous], 2017, ARXIV170200288
[3]
[Anonymous], 1990, Neurocomputing: Algorithms, architectures and applications
[4]
Barboriak D., 2015, DATA RIDER NEURO MRI, DOI 10.7937/K9/TCIA2015VOSN3HN1
[5]
BENOU A, 2016, NOIS CON TRAST ENH, V8, P95
[6]
PHARMACOKINETIC PARAMETERS IN CNS GD-DTPA ENHANCED MR IMAGING [J].
BRIX, G ;
SEMMLER, W ;
PORT, R ;
SCHAD, LR ;
LAYER, G ;
LORENZ, WJ .
JOURNAL OF COMPUTER ASSISTED TOMOGRAPHY, 1991, 15 (04) :621-628
[7]
A non-local algorithm for image denoising [J].
Buades, A ;
Coll, B ;
Morel, JM .
2005 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, VOL 2, PROCEEDINGS, 2005, :60-65
[8]
Nonlocal image and movie denoising [J].
Buades, Antoni ;
Coll, Bartomeu ;
Morel, Jean-Michel .
INTERNATIONAL JOURNAL OF COMPUTER VISION, 2008, 76 (02) :123-139
[9]
Analyzing the blood-brain barrier: The benefits of medical imaging in research and clinical practice [J].
Chassidim, Yoash ;
Vazana, Udi ;
Prager, Ofer ;
Veksler, Ronel ;
Bar-Klein, Guy ;
Schoknecht, Karl ;
Fassler, Michael ;
Lublinsky, Svetlana ;
Shelef, Ilan .
SEMINARS IN CELL & DEVELOPMENTAL BIOLOGY, 2015, 38 :43-52
[10]
Patch-Based Near-Optimal Image Denoising [J].
Chatterjee, Priyam ;
Milanfar, Peyman .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2012, 21 (04) :1635-1649