Enabling Augmented Segmentation and Registration in Ultrasound-Guided Spinal Surgery via Realistic Ultrasound Synthesis from Diagnostic CT Volume

This paper aims to tackle the issues on unavailable or insufficient clinical US data and meaningful annotation to enable bone segmentation and registration for US-guided spinal surgery. While the US is not a standard paradigm for spinal surgery, the scarcity of intra-operative clinical US data is an insurmountable bottleneck in training a neural network. Moreover, due to the characteristics of US imaging, it is difficult to clearly annotate bone surfaces which causes the trained neural network missing its attention to the details. Hence, we propose an In silico bone US simulation framework that synthesizes realistic US images from diagnostic CT volume. Afterward, using these simulated bone US we train a lightweight vision transformer model that can achieve accurate and on-the-fly bone segmentation for spinal sonography. In the validation experiments, the realistic US simulation was conducted by deriving from diagnostic spinal CT volume to facilitate a radiation-free US-guided pedicle screw placement procedure. When it is employed for training bone segmentation task, the Chamfer distance achieves 0.599mm; when it is applied for CT-US registration, the associated bone segmentation accuracy achieves 0.93 in Dice, and the registration accuracy based on the segmented point cloud is 0.13~3.37mm in a complication-free manner. While bone US images exhibit strong echoes at the medium interface, it may enable the model indistinguishable between thin interfaces and bone surfaces by simply relying on small neighborhood information. To overcome these shortcomings, we propose to utilize a Long-range Contrast Learning Module to fully explore the Long-range Contrast between the candidates and their surrounding pixels.

[1]  Jian Sun,et al.  Scaling Up Your Kernels to 31×31: Revisiting Large Kernel Design in CNNs , 2022, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Mohammad Rastegari,et al.  MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer , 2021, ICLR.

[3]  M. M. A. Hashem,et al.  CNL-UNet: A novel lightweight deep learning architecture for multimodal biomedical image segmentation with false output suppression , 2021, Biomed. Signal Process. Control..

[4]  Longin Jan Latecki,et al.  SGUNET: Semantic Guided UNET For Thyroid Nodule Segmentation , 2021, 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI).

[5]  Hongen Liao,et al.  Autonomic Robotic Ultrasound Imaging System Based on Reinforcement Learning , 2021, IEEE Transactions on Biomedical Engineering.

[6]  Bo Peng,et al.  A Real-time Ultrasound Simulator Using Monte-Carlo Path Tracing in Conjunction with Optix Engine , 2020, 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC).

[7]  Debdoot Sheet,et al.  Simulating patho-realistic ultrasound images using deep generative networks with adversarial learning , 2017, 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018).

[8]  Dean C. Barratt,et al.  Freehand Ultrasound Image Simulation with Spatially-Conditioned Generative Adversarial Networks , 2017, CMMI/RAMBO/SWITCH@MICCAI.

[9]  I. Hacihaliloglu Ultrasound imaging and segmentation of bone surfaces: A review. , 2017, Technology.

[10]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[11]  Yongtian Wang,et al.  Fast and Automatic Ultrasound Simulation from CT Images , 2013, Comput. Math. Methods Medicine.

[12]  Andras Lasso,et al.  Improving N-wire phantom-based freehand ultrasound calibration , 2013, International Journal of Computer Assisted Radiology and Surgery.

[13]  Adam Piórkowski,et al.  The Transesophageal Echocardiography Simulator Based on Computed Tomography Images , 2013, IEEE Transactions on Biomedical Engineering.

[14]  Alistair P. Rendell,et al.  Modeling nonlinear ultrasound propagation in heterogeneous media with power law absorption using a k-space pseudospectral method. , 2012, The Journal of the Acoustical Society of America.

[15]  Mohammad Talebi,et al.  Medical ultrasound image segmentation using genetic active contour , 2011 .

[16]  Xianglong Tang,et al.  Probability density difference-based active contour for ultrasound image segmentation , 2010, Pattern Recognit..

[17]  Nassir Navab,et al.  Visualization and GPU-accelerated simulation of medical ultrasound from CT images , 2009, Comput. Methods Programs Biomed..

[18]  Nassir Navab,et al.  Automatic CT-ultrasound registration for diagnostic imaging and image-guided intervention , 2008, Medical Image Anal..

[19]  Peter R Hoskins,et al.  Physical properties of tissues relevant to arterial ultrasound imaging and blood velocity measurement. , 2007, Ultrasound in medicine & biology.

[20]  B T Cox,et al.  Fast calculation of pulsed photoacoustic fields in fluids using k-space methods. , 2005, The Journal of the Acoustical Society of America.

[21]  Kay Raum,et al.  Frequency and resolution dependence of the anisotropic impedance estimation in cortical bone using time-resolved scanning acoustic microscopy. , 2004, Journal of biomedical materials research. Part A.

[22]  Jørgen Jensen,et al.  Simulation of advanced ultrasound systems using Field II , 2004, 2004 2nd IEEE International Symposium on Biomedical Imaging: Nano to Macro (IEEE Cat No. 04EX821).

[23]  B. Treeby,et al.  Modelling nonlinear ultrasound propagation in absorbing media using the k-Wave toolbox: experimental validation , 2012, 2012 IEEE International Ultrasonics Symposium.

[24]  Jürgen Hesser,et al.  Simulation of Dynamic Ultrasound Based on CT Models for Medical Education , 2008, MMVR.

[25]  J. Arendt Paper presented at the 10th Nordic-Baltic Conference on Biomedical Imaging: Field: A Program for Simulating Ultrasound Systems , 1996 .

[26]  K. Parker,et al.  Absorption and attenuation in soft tissues. II. Experimental results , 1988, IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control.