Artifact Suppression for Passive Cavitation Imaging Using U-Net CNNs with Uncertainty Quantification

The clinical applications of chemotherapy treatments for brain tumors are limited by the blood-brain barrier (BBB) which blocks drugs from reaching the tumors. Magnetic resonance imaging (MRI)-guided focused ultrasound (FUS) is applied to the brain to create micro-bubbles in the BBB for opening the BBB temporarily for targeted drug delivery. Creation of these microbubbles is monitored using passive cavitation imaging (PCI). Generally, PCI utilizes delay-and-sum (DAS) beam-forming, which suffers from a space-varying point spread function and high sidelobes both of which introduce image artifacts. To address this issue and to better monitor the BBB opening, a deep learning denoising approach is proposed based on a U-Net convolutional neural networks (CNN) architecture. The U-Net is trained with pairs of images, with DAS passive cavitation images at the input layer, and ground true noise-free images at the output layer. In the training phase, the network fits the mapping from the DAS images to the noise-free images. A Monte Carlo (MC) dropout approach is used to quantify uncertainty in de-noised images, giving information that can be further used to suppress artifacts. It is shown in this work that the resulting model can generate de-noised DAS images, and that uncertainty quantification in U-net output can be used to further improve artifact suppression.

[1]  Miklós Gyöngy,et al.  Passive Spatial Mapping of Inertial Cavitation During HIFU Exposure , 2010, IEEE Transactions on Biomedical Engineering.

[2]  Michael Unser,et al.  Deep Convolutional Neural Network for Inverse Problems in Imaging , 2016, IEEE Transactions on Image Processing.

[3]  Chanikarn Power,et al.  Closed-loop control of targeted ultrasound drug delivery across the blood–brain/tumor barriers in a rat glioma model , 2017, Proceedings of the National Academy of Sciences.

[4]  Sergey Ioffe,et al.  Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning , 2016, AAAI.

[5]  Jay Jagannathan,et al.  HIGH‐INTENSITY FOCUSED ULTRASOUND SURGERY OF THE BRAIN: PART 1—A HISTORICAL PERSPECTIVE WITH MODERN APPLICATIONS , 2009, Neurosurgery.

[6]  Vasant A Salgaonkar,et al.  Passive cavitation imaging with ultrasound arrays. , 2009, The Journal of the Acoustical Society of America.

[7]  Gregory T. Clement,et al.  Clinical applications of focused ultrasound—The brain , 2007, International journal of hyperthermia : the official journal of European Society for Hyperthermic Oncology, North American Hyperthermia Group.

[8]  Arthur D. Yaghjian,et al.  Plane-wave theory of time-domain fields : near-field scanning applications , 1999 .

[9]  D. Rawlins,et al.  The point‐spread function of a confocal microscope: its measurement and use in deconvolution of 3‐D data , 1991 .

[10]  Stephan Antholzer,et al.  Deep learning for photoacoustic tomography from sparse data , 2017, Inverse problems in science and engineering.

[11]  Ronald A. Roy,et al.  Applications of Acoustics and Cavitation to Noninvasive Therapy and Drug Delivery , 2008 .

[12]  Christopher Joseph Pal,et al.  The Importance of Skip Connections in Biomedical Image Segmentation , 2016, LABELS/DLMIA@MICCAI.

[13]  A. E. Miller,et al.  A NEW METHOD FOR THE GENERATION AND USE OF FOCUSED ULTRASOUND IN EXPERIMENTAL BIOLOGY , 1942, The Journal of general physiology.

[14]  Robert H. Mellen,et al.  Ultrasonic Spectrum of Cavitation Noise in Water , 1954 .

[15]  Penny Probert Smith,et al.  Passive acoustic mapping utilizing optimal beamforming in ultrasound therapy monitoring. , 2015, The Journal of the Acoustical Society of America.

[16]  Miklós Gyöngy,et al.  Passive cavitation mapping for localization and tracking of bubble dynamics. , 2010, The Journal of the Acoustical Society of America.

[17]  Zoubin Ghahramani,et al.  Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning , 2015, ICML.

[18]  P. Cochat,et al.  Et al , 2008, Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.