Bi-Directional Semi-Supervised Training of Convolutional Neural Networks for Ultrasound Elastography Displacement Estimation

Tehrani, Ali K. Z.; Sharifzadeh, Mostafa; Boctor, Emad M.; Rivaz, Hassan

doi:10.1109/tuffc.2022.3147097

Cited by 14 publications

(11 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where a and l represent the axial and lateral directions, respectively. The parameter β depends on the ratio of the spatial distance between two samples in axial and lateral directions, and is set it to 0.1 similar to [15]. Finally, PICTURE loss is obtained as:…”

Section: Physically Inspired Constraint For Unsupervised Regularized ...mentioning

confidence: 99%

“…Let I 1 , I 2 ∈ R 3×w×h be the pre-compression, and post-compression US data each has the width h, height w and the 3 channels of RF data, the imaginary part of the analytic signal and the envelope of RF data, respectively (similar to [15]). The data loss in unsupervised training is the photometric loss obtained by comparing I 1 and warped I 2 ( Ĩ2 ) using bi-linear warping by the displacement W which can be defined as [15,13]:…”

Section: Unsupervised Trainingmentioning

confidence: 99%

“…Therefore, regularization of the displacement has been employed. We adopt the regularization method of [15] where the first-and second-order derivatives of displacement (strains and it's first-order derivative) are employed:…”

Section: Unsupervised Trainingmentioning

confidence: 99%

“…The < . > denotes the mean operation, and the hyperparameters are tuned similar to [15]. We propose to add the PICTURE loss as the regularization constraint to improve the lateral displacement quality.…”

Section: Unsupervised Trainingmentioning

confidence: 99%

“…They have been successfully adopted for USE by modifying the architectures to handle high-frequency radio-frequency (RF) data [14,6]. Unsupervised and semi-supervised techniques have also been employed to train the networks using real US data without requiring the ground truth displacements [13,1,2,15]. They employed prior knowledge of displacement continuity in the form of the first-and second-derivative regularizers.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Physically Inspired Constraint for Unsupervised Regularized Ultrasound Elastography

Tehrani¹,

Rivaz²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Displacement estimation is a critical step of virtually all Ultrasound Elastography (USE) techniques. Two main features make this task unique compared to the general optical flow problem: the highfrequency nature of ultrasound radio-frequency (RF) data and the governing laws of physics on the displacement field. Recently, the architecture of the optical flow networks has been modified to be able to use RF data. Also, semi-supervised and unsupervised techniques have been employed for USE by considering prior knowledge of displacement continuity in the form of the first-and second-derivative regularizers. Despite these attempts, no work has considered the tissue compression pattern, and displacements in axial and lateral directions have been assumed to be independent. However, tissue motion pattern is governed by laws of physics in USE, rendering the axial and the lateral displacements highly correlated. In this paper, we propose Physically Inspired ConsTraint for Unsupervised Regularized Elastography (PICTURE), where we impose constraints on the Poisson's ratio to improve lateral displacement estimates. Experiments on phantom and in vivo data show that PICTURE substantially improves the quality of the lateral displacement estimation.

show abstract

Section: Physically Inspired Constraint For Unsupervised Regularized ...mentioning

confidence: 99%

Section: Unsupervised Trainingmentioning

confidence: 99%

Section: Unsupervised Trainingmentioning

confidence: 99%

Section: Unsupervised Trainingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Physically Inspired Constraint for Unsupervised Regularized Ultrasound Elastography

Tehrani¹,

Rivaz²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

Alternating direction method of multipliers for displacement estimation in ultrasound strain elastography

Ashikuzzaman,

Peng,

Jiang

et al. 2023

Medical Physics

Self Cite

View full text Add to dashboard Cite

BackgroundUltrasound strain imaging, which delineates mechanical properties to detect tissue abnormalities, involves estimating the time delay between two radio‐frequency (RF) frames collected before and after tissue deformation. The existing regularized optimization‐based time‐delay estimation (TDE) techniques suffer from at least one of the following drawbacks: (1) The regularizer is not aligned with the tissue deformation physics due to taking only the first‐order displacement derivative into account; (2) The L2‐norm of the displacement derivatives, which oversmooths the estimated time‐delay, is utilized as the regularizer; (3) The modulus function defined mathematically should be approximated by a smooth function to facilitate the optimization of L1‐norm.PurposeOur purpose is to develop a novel TDE technique that resolves the aforementioned shortcomings of the existing algorithms.MethodsHerein, we propose employing the alternating direction method of multipliers (ADMM) for optimizing a novel cost function consisting of L2‐norm data fidelity term and L1‐norm first‐ and second‐order spatial continuity terms. ADMM empowers the proposed algorithm to use different techniques for optimizing different parts of the cost function and obtain high‐contrast strain images with smooth backgrounds and sharp boundaries. We name our technique ADMM for totaL variaTion RegUlarIzation in ultrasound STrain imaging (ALTRUIST). ALTRUIST's efficacy is quantified using absolute error (AE), Structural SIMilarity (SSIM), signal‐to‐noise ratio (SNR), contrast‐to‐noise ratio (CNR), and strain ratio (SR) with respect to GLUE, OVERWIND, and L1‐SOUL, three recently published energy‐based techniques, and UMEN‐Net, a state‐of‐the‐art deep learning‐based algorithm. Analysis of variance (ANOVA)‐led multiple comparison tests and paired t‐tests at 5% overall significance level were conducted to assess the statistical significance of our findings. The Bonferroni correction was taken into account in all statistical tests. Two simulated layer phantoms, three simulated resolution phantoms, one hard‐inclusion simulated phantom, one multi‐inclusion simulated phantom, one experimental breast phantom, and three in vivo liver cancer datasets have been used for validation experiments. We have published the ALTRUIST code at http://code.sonography.ai.ResultsALTRUIST substantially outperforms the four state‐of‐the‐art benchmarks in all validation experiments, both qualitatively and quantitatively. ALTRUIST yields up to , , and SNR improvements and , , and CNR improvements over L1‐SOUL, its closest competitor, for simulated, phantom, and in vivo liver cancer datasets, respectively, where the asterisk (*) indicates statistical significance. In addition, ANOVA‐led multiple comparison tests and paired t‐tests indicate that ALTRUIST generally achieves statistically significant improvements over GLUE, UMEN‐Net, OVERWIND, and L1‐SOUL in terms of AE, SSIM map, SNR, and CNR.ConclusionsA novel ultrasonic displacement tracking algorithm named ALTRUIST has been developed. The principal novelty of ALTRUIST is incorporating ADMM for optimizing an L1‐norm regularization‐based cost function. ALTRUIST exhibits promising performance in simulation, phantom, and in vivo experiments.

show abstract

Physically Inspired Constraint for Unsupervised Regularized Ultrasound Elastography

Tehrani

Rivaz

2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Bi-Directional Semi-Supervised Training of Convolutional Neural Networks for Ultrasound Elastography Displacement Estimation

Cited by 14 publications

References 36 publications

Physically Inspired Constraint for Unsupervised Regularized Ultrasound Elastography

Physically Inspired Constraint for Unsupervised Regularized Ultrasound Elastography

Alternating direction method of multipliers for displacement estimation in ultrasound strain elastography

Physically Inspired Constraint for Unsupervised Regularized Ultrasound Elastography

Contact Info

Product

Resources

About