Quantitative Ultrasound & Radar Imaging

Introduction

We focus on new radar (RD) and ultrasound (US) quantitative biomedical imaging algorithms and systems that are expected to quickly and accurately reconstruct the electrical and acoustic parameters of tissues and organs, realizing structural and functional imaging with high efficiency. Our solution is to develop efficient, high performance (yet interpretable), low training set reliant deep learning algorithms and network structures for RD and US quantitative imaging. In addition, specific hardware systems are built for data collection and algorithm validation.

Nonlinear US full-waveform inversion

We developed the nonlinear full waveform inversion (FWI) algorithm, which is based on representing the wave equation as an RNN, with an architecture determined by the physical acoustic model. This representation enables to apply optimization algorithms borrowed from the deep-learning toolbox and paves the way to very efficient implementation.
In the illustration of the simulated medium figure below, water is represented in white background and the left and right objects are fat and liver, respectively. The tissues and the water have different speed of sound (SoS), density, attenuation, and nonlinearity, while the transducer array is located at the top. In the reconstruction of simulated medium figure, the first row shows the ground truth (GT) values of the simulated medium. The second row shows the initial values used as input to the inverse algorithm. In the third row presents the obtained reconstructions using NWI. We evaluated the reconstruction error using the normalized RMSE metric, demonstrating the high reconstruction performance of our method.

$Screenshot 2026-07-29 130025$

https://youtu.be/BSX6w3lAabo

To facilitate FWI in clinical route examination, we further develop a frequency-domain, total variation (TV)-regularized FWI framework tailored for single-sided linear ultrasound array, which enables quantitative reconstruction of SoS maps using standard clinical probes. To address the severe ill-posedness and computational challenges in this setup, efficient forward modeling, fast gradient evaluation, ADMM-based optimization, and multi-GPU parallelization are integrated into the inversion framework. Numerical experiments in a thyroid cyst imaging scenario demonstrate that the proposed method reconstructs the SoS of both simple (fluid-filled) and solid cysts with improved visual and quantitative performance compared to conventional FWI. We also demonstrated SoS recovery of tissue-mimicking phantoms based on our framework — including materials spanning the range from soft tissue to cartilage-like SoS and stiffer bone tumors— using a Telemed phased array probe, demonstrating feasibility for real-world clinical deployment.

$deployment$

$Screenshot 2026-07-29 130341$

$Screenshot 2026-07-29 130519$

Quantitative model-based deep neural network for US and RD imaging

We introduce MB-QRUS, which stands for Model-Based Quantitative Radar and US, which is based on an unfolding mechanism of with learned gradients according to a U-Net based block. We use the residuals between the measured channel data and the predicted channel data, according to physical property estimation and the physical model of wave propagation, to learn the gradients which are often used to update the physical properties estimation. We also introduce a new time-domain and tensor representation of the input measured channel which captures the spatial representation of the channel data.

$Picture10$

$Picture11$

We further proposed DUFWI, deep unfolding-based approach for FWI, for quantitative US imaging. To meet the computational demands of our time-domain method, we adopt block-by-block training, allowing for the reuse of computed gradients in each epoch. This method efficiently trains the deep neural network at each iteration of the deep unfolded algorithm, enabling the use of larger simulated domains during training—a key requirement for real-world applications. Both simulation and hardware experiments on the training dataset and out-of-distribution examples show that our method outperforms classical FWI and MB-QRUS in reconstruction quality under noisy conditions while maintaining a high level of efficiency.

$Screenshot 2026-07-29 130602$

$Picture13$

Quantitative inverse imaging with generative models

We propose L-DPnP, a latent diffusion plug-and-play method for quantitative inverse imaging in both RD and US applications. This method samples from the posterior distribution of electrical or acoustic property maps given array measurements. It integrates a data consistency sampler to generate measurement-consistent samples, utilizing a latent diffusion model combined with a variational autoencoder as an expressive prior. Our method is first validated on a two-dimensional microwave dataset simulated from human head models with and without hemorrhagic or ischemic strokes. Compared to existing methods, L-DPnP achieves the lowest model misfit (0.067 vs. ≥ 0.085) and the highest structural similarity index measure (SSIM) (0.936 vs. ≤ 0.906). Furthermore, L-DPnP is validated in quantitative US imaging for edema detection, where it outperforms both DUFWI and classical FWI. Specifically, it yields a higher SSIM (0.9869, compared to 0.9846 for DUFWI and 0.9116 for classical FWI) and a smaller model misfit (-32.28 dB, compared to -30.81 dB and -23.65 dB, respectively).

$Screenshot 2026-07-29 130756$

$Screenshot 2026-07-19 150919$

$Picture15$

References:

Shultzman, Avner, and Yonina C. Eldar. "Nonlinear waveform inversion for quantitative ultrasound." IEEE Transactions on Computational Imaging 8 (2022): 893-904.
Guo, Rui, Tianyao Huang, Maokun Li, Haiyang Zhang, and Yonina C. Eldar. "Physics-embedded machine learning for electromagnetic data imaging: Examining three types of data-driven imaging methods." IEEE Signal Processing Magazine 40, no. 2 (2023): 18-31.
Sharon, Tom, and Yonina C. Eldar. "Real-time model-based quantitative ultrasound and radar." IEEE Transactions on Computational Imaging 10 (2024):1175-1190.
Cohen, Niv, Yhonatan Kvich, Rui Guo, and Yonina C. Eldar. "Deep Unfolding of Full Waveform Inversion for Quantitative Ultrasound Imaging." In ICASSP 2025-2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 1-5. IEEE, 2025.
Guo, Rui, Yhonatan Kvich, Yi Zhang, Tianyao Huang, Maokun Li, and Yonina C. Eldar. "Posterior Sampling with Latent Diffusion for Microwave Brain Imaging." In 2025 47th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 1-7. IEEE, 2025.
Guo, Rui, Yi Zhang, Yhonatan Kvich, Tianyao Huang, Maokun Li, and Yonina C. Eldar., Plug-and-Play Latent Diffusion for Electromagnetic Inverse Scattering with Application to Brain Imaging (https://arxiv.org/pdf/2509.04860).
Guo, Rui, Ditza Auerbach, Yonina C. Eldar. TV-Regularized Frequency-Domain Full-Waveform Inversion for Single-Sided Linear Ultrasound Array Data, 48th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 1-7. IEEE, 2026.
Zhang, Ruizhi, Rui Guo, Yi Zhang, Yhonatan Kvich, Yonina C. Eldar, Plug-and-Play Latent Diffusion for Quantitative Ultrasound Imaging, IEEE Sensor Array and Multichannel Signal Processing Workshop 2026.