Comparison of texture synthesis methods for content generation in ultrasound simulation for training

@inproceedings{Mattausch2017ComparisonOT,
  title={Comparison of texture synthesis methods for content generation in ultrasound simulation for training},
  author={Oliver Mattausch and Elizabeth Ren and Michael Bajka and Kenneth Vanhoey and Orcun Goksel},
  booktitle={Medical Imaging},
  year={2017}
}
Navigation and interpretation of ultrasound (US) images require substantial expertise, the training of which can be aided by virtual-reality simulators. However, a major challenge in creating plausible simulated US images is the generation of realistic ultrasound speckle. Since typical ultrasound speckle exhibits many properties of Markov Random Fields, it is conceivable to use texture synthesis for generating plausible US appearance. In this work, we investigate popular classes of texture… 

Ultrasound simulation with deformable and patient-specific scatterer maps

TLDR
This work realistically simulate US texture deformations in the scatterer domain via back-projection of ray segments into a nominal state before sampling during simulation runtime, and estimates scatterers maps from background in vivo images using a pretrained generative adversarial network.

Framework for Fusion of Data- and Model-Based Approaches for Ultrasound Simulation

TLDR
The proposed hybrid simulation method is tested by replacing embryos within gestational sacs by ray-traced embryos, and by simulating an ectoptic pregnancy, to cover the range of rare findings or particular clinical tasks with known ground truth geometric models.

Sketch guided and progressive growing GAN for realistic and editable ultrasound image synthesis

TLDR
The first work that can synthesize realistic B-mode US images with high-resolution and customized texture editing features is presented, and a feature loss is proposed to minimize the difference of high-level features between the generated and real images.

Interactive Translation in Echocardiography Training System With Enhanced Cycle-GAN

TLDR
Perceptual loss is introduced to enhance the quality of synthetic ultrasound texture, while Cycle-GAN translates between two modalities, and the proposed method is feasible in interactive translation, and it is superior over cycles for ultrasound image synthesis.

Thoracic MRI Emulation Through Texture Synthesis and Ground Truth Manipulation

TLDR
The results indicate that generator architectures with residual learning blocks are efficient for the image synthesis and that single-channel annotations convey information with more efficacy than the one-hot encoded ones.

Image-Based Reconstruction of Tissue Scatterers Using Beam Steering for Ultrasound Simulation

TLDR
A principled approach is introduced to estimate (reconstruct) a scatterer map from images, by solving the inverse-problem of ultrasound speckle formation, such that images from arbitrary view angles and transducer settings can be generated from those scatterers later in simulations.

References

SHOWING 1-10 OF 31 REFERENCES

Texture optimization for example-based synthesis

TLDR
A Markov Random Field-based similarity metric for measuring the quality of synthesized texture with respect to a given input sample is defined, which allows the synthesis problem as minimization of an energy function, which is optimized using an Expectation Maximization (EM)-like algorithm.

On pixel-based texture synthesis by non-parametric sampling

Self Tuning Texture Optimization

TLDR
This work presents a new general‐purpose and fully automatic self‐tuning non‐parametric texture synthesis method that extends Texture Optimization by introducing several key improvements that result in superior synthesis ability.

State of the Art in Example-based Texture Synthesis

TLDR
This report aims to provide a tutorial that is easy to follow for readers who are not already familiar with the subject, make a comprehensive survey and comparisons of different methods, and sketch a vision for future work that can help motivate and guide readers that are interested in texture synthesis research.

B-Mode Ultrasound Image Simulation in Deformable 3-D Medium

TLDR
An algorithm for fast image synthesis inside deformed volumes and an ultrasound examination system was implemented to demonstrate that real-time image synthesis with the proposed technique can be successfully integrated into a haptic simulation.

Pyramid-based texture analysis/synthesis

TLDR
This paper describes a method for synthesizing images that match the texture appearance of a given digitized sample, based on a model of human texture perception, and has potential to be a practically useful tool for graphics applications.

Monte-Carlo Ray-Tracing for Realistic Interactive Ultrasound Simulation

TLDR
Monte-Carlo ray-tracing for ultrasound enables the realistic simulation of ultrasound interactions such as soft shadows and fuzzy reflections and discusses how to properly weight the contribution of each ray path in order to simulate the behavior of a beamformed ultrasound signal.

The Synthesizability of Texture Examples

TLDR
This work is the first attempt to quantify this image property, and it is found that texture synthesizability can be learned and predicted and used to trim images to parts that are more synthesizable.

Synthesizing natural textures

TLDR
The algorithm is extended to allow direct user input for interactive control over the texture synthesis process, which allows the user to indicate large-scale properties of the texture appearance using a standard painting-style interface, and to choose among various candidate textures the algorithm can create by performing different number of iterations.

A Serious Game for Learning Ultrasound-Guided Needle Placement Skills

TLDR
A serious game to teach the skills involved in ultrasound-guided needle placement training and introduces novel techniques for realistic simulation and integrates game elements for active and effective learning.