Iterative Pulse–Echo Tomography for Ultrasonic Image Correction

Yu-chen Zengqiu, Wentao Wu, Ling Xiao, Erlei Zhou, Zheng Cao, Jiadong Hua, Yue WangSensors 2024 · DOI

TL;DR

This paper introduces an "iterative CUTE" algorithm to improve ultrasound image quality by more accurately estimating the speed of sound (SoS) within tissues. By iteratively refining SoS estimates, the method significantly reduces image distortions caused by acoustic aberrations, outperforming traditional single-pass techniques.

What Problem Does This Paper Address?

Medical ultrasound imaging is a powerful diagnostic tool, but its quality is often degraded by acoustic aberrations. These aberrations occur because ultrasound waves travel at different speeds through various tissues, a phenomenon not accounted for in standard imaging which assumes a uniform speed of sound (around 1540 m/s). This mismatch distorts the ultrasound wavefront, leading to blurry images with reduced resolution and contrast. While methods like Computed Ultrasound Tomography in Echo Mode (CUTE) exist to estimate the local speed of sound (SoS) for correction, the traditional single-pass CUTE approach struggles with accuracy and robustness, especially in complex tissue structures.

Methods

The core innovation is an iterative CUTE algorithm. Unlike the traditional single-pass CUTE, which estimates SoS once, the iterative approach refines the SoS estimate through multiple correction steps.

Here's a breakdown:

  • Forward Model: The paper starts with a forward model that relates the local SoS distribution to the phase shift observed in ultrasound echo signals. This model is discretized for numerical computation.
  • Inversion: To estimate the SoS, the phase shift measurements are "inverted" using a regularized pseudo-inverse (Tikhonov regularization) to obtain slowness deviation estimates.
  • Iterative Correction: This is the key novelty. Instead of using a fixed presumed SoS for beam-forming and phase-shift extraction, the algorithm uses the current estimate of the local SoS from the previous iteration. This allows the phase shift to reflect the error between the current SoS estimate and the true SoS. This error is then used to correct the SoS estimate in each iteration. The process repeats until the change in the SoS estimate falls below a threshold.
  • Coherence Factor (CF): The authors propose using the Coherence Factor (CF) as a more relevant metric than traditional Root Mean Square Error (RMSE) for evaluating the quality of SoS estimates for imaging correction. CF directly assesses how well the SoS estimate improves image focus.

Datasets

The paper uses two types of datasets for evaluation:

  • Simulated Phantoms: Phantoms with inclusions of various shapes (cylindrical, oval, irregular strip-like) and different SoS values were simulated. These were used to test the algorithm's accuracy and robustness.
  • Agar Phantom Experiment: A physical agar phantom with a shredded ham strip (higher SoS inclusion) was created and imaged. This provided a real-world validation of the iterative CUTE algorithm.

Key Insights and Ideas

  • Single-pass CUTE limitations: The paper highlights that single-pass SoS estimation methods, while practical, suffer from poor accuracy and robustness due to factors like regularization choices and inherent noise.
  • Iterative refinement is crucial: The iterative nature of the proposed algorithm is key to overcoming the limitations of single-pass methods. By repeatedly correcting the SoS estimate, the algorithm can better balance the trade-off between suppressing interference (from strong regularization) and achieving sufficient correction (from weak regularization).
  • CF is a better metric for imaging correction: Traditional metrics like RMSE don't directly correlate with image quality improvements. The Coherence Factor (CF) provides a more direct measure of how well an SoS estimate translates to better image focus and clarity.
  • Robustness to SoS assumption: The iterative approach demonstrates robustness even when the initial presumed SoS value has a significant discrepancy from the true value.

Key Results

  • Simulation: The iterative CUTE approach enhanced the average Coherence Factor (CF) for SoS estimates by up to 18.2% compared to single-pass methods.
  • Phantom Experiments: Imaging corrected with SoS estimates from iterative CUTE reduced the Array Performance Index (API) by up to 40% compared to traditional methods, indicating significantly improved image quality.
  • Shape Robustness: The iterative algorithm showed improved reconstruction of irregularly shaped inclusions, which are challenging for single-pass methods.
  • Regularization Trade-off: The iterative method effectively navigates the regularization challenges, achieving better results across a range of regularization strengths compared to single-pass methods.

Limitations and Future Work

  • Theoretical Proof of Convergence: While the iterative algorithm demonstrated convergence in simulations and phantom experiments, a theoretical proof of this convergence is still needed. This is identified as a focus for future work.
  • Computational Cost: Although not explicitly detailed as a limitation, iterative algorithms generally have higher computational costs than single-pass methods. The paper focuses on the quality improvement, implying a trade-off with computation time.

Summary updated

Abstract

Acoustic aberration, caused by the uneven distribution of tissue speed-of-sound (SoS), significantly reduces the quality of ultrasound imaging. An important approach to mitigate this issue is imaging correction based on local SoS estimation. Computed ultrasound tomography in echo mode (CUTE) is an SoS estimation method that utilizes phase-shift information from ultrasound pulse-echo signals, offering both practical utility and computational efficiency. However, the traditional single-pass CUTE often suffers from poor accuracy and robustness. In this paper, an advanced approach known as iterative CUTE is introduced, which refines SoS estimates through iterative correction of errors and noise, addressing the limitations of traditional single-pass methods. It is argued that traditional precision indicators like root mean square error (RMSE) fall short in adequately reflecting the quality of SoS estimates for imaging correction, and coherence factor (CF) is proposed as a more indicative metric. Performance validation of the iterative CUTE algorithm was conducted using a simulation and agar phantom experiment. The results indicated that the iterative CUTE approach surpasses the single-pass approach, enhancing the average CF for SoS estimates by up to 18.2%. In phantom experiments, imaging corrected with SoS estimates from iterative CUTE reduced the Array Performance Index (API) by up to 40% compared to traditional methods.
BibTeX
@inproceedings{zengqiu2024iterative,
  author = {Yu-chen Zengqiu and Wentao Wu and Ling Xiao and Erlei Zhou and Zheng Cao and Jiadong Hua and Yue Wang},
  title = {{Iterative Pulse–Echo Tomography for Ultrasonic Image Correction}},
  booktitle = {Sensors},
  year = {2024},
  doi = {10.3390/s24061895},
}
Figure