Evaluating HIFU-mediated local drug release using thermal strain imaging: Phantom and preliminary in-vivo studies

Dense speed-of-sound shift imaging for ultrasonic thermometry

Objective. Develop a dense algorithm for calculating the speed-of-sound shift between consecutive acoustic acquisitions as a noninvasive means to evaluating temperature change during thermal ablation.Methods. An algorithm for dense speed-of-sound shift imaging (DSI) was developed to simultaneously incorporate information from the entire field of view using a combination of dense optical flow and inverse problem regularization, thus speeding up the calculation and introducing spatial agreement between pixels natively. Thermal ablation monitoring consisted of two main steps: pixel shift tracking using Farneback optical flow, and mathematical modeling of the relationship between the pixel displacement and temperature change as an inverse problem to find the speed-of-sound shift. A calibration constant translates from speed-of-sound shift to temperature change. The method performance was tested inex vivosamples and compared to standard thermal strain imaging (TSI) methods.Main results. Thermal ablation at a frequency of 2 MHz was applied to an agarose phantom that created a speed-of-sound shift measured by an L12-5 imaging transducer. A focal spot was reconstructed by solving the inverse problem. Next, a thermocouple measured the temperature rise during thermal ablation ofex vivochicken breast to calibrate the setup. Temperature changes between 3 °C and 15 °C was measured with high thermometry precision of less than 2 °C error for temperature changes as low as 8 °C. The DSI method outperformed standard TSI in both spatial coherence and runtime in high-intensity focused ultrasound-induced hyperthermia.Significance. Dense ultrasonic speed-of-sound shift imaging can successfully monitor the speed-of-sound shift introduced by thermal ablation. This technique is faster and more robust than current methods, and therefore can be used as a noninvasive, real time and cost-effective thermometry method, with high clinical applicability.

Lateral Strain Imaging Using Self-Supervised and Physically Inspired Constraints in Unsupervised Regularized Elastography

Convolutional Neural Networks (CNN) have shown promising results for displacement estimation in UltraSound Elastography (USE). Many modifications have been proposed to improve the displacement estimation of CNNs for USE in the axial direction. However, the lateral strain, which is essential in several downstream tasks such as the inverse problem of elasticity imaging, remains a challenge. The lateral strain estimation is complicated since the motion and the sampling frequency in this direction are substantially lower than the axial one, and a lack of carrier signal in this direction. In computer vision applications, the axial and the lateral motions are independent. In contrast, the tissue motion pattern in USE is governed by laws of physics which link the axial and lateral displacements. In this paper, inspired by Hooke's law, we, first propose Physically Inspired ConsTraint for Unsupervised Regularized Elastography (PICTURE), where we impose a constraint on the Effective Poisson's ratio (EPR) to improve the lateral strain estimation. In the next step, we propose self-supervised PICTURE (sPICTURE) to further enhance the strain image estimation. Extensive experiments on simulation, experimental phantom and in vivo data demonstrate that the proposed methods estimate accurate axial and lateral strain maps.

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

The performance of ultrasound elastography (USE) heavily depends on the accuracy of displacement estimation. Recently, convolutional neural networks (CNNs) have shown promising performance in optical flow estimation and have been adopted for USE displacement estimation. Networks trained on computer vision images are not optimized for USE displacement estimation since there is a large gap between the computer vision images and the high-frequency radio frequency (RF) ultrasound data. Many researchers tried to adopt the optical flow CNNs to USE by applying transfer learning to improve the performance of CNNs for USE. However, the ground-truth displacement in real ultrasound data is unknown, and simulated data exhibit a domain shift compared to the real data and are also computationally expensive to generate. To resolve this issue, semisupervised methods have been proposed in which the networks pretrained on computer vision images are fine-tuned using real ultrasound data. In this article, we employ a semisupervised method by exploiting the first- and second-order derivatives of the displacement field for regularization. We also modify the network structure to estimate both forward and backward displacements and propose to use consistency between the forward and backward strains as an additional regularizer to further enhance the performance. We validate our method using several experimental phantom and in vivo data. We also show that the network fine-tuned by our proposed method using experimental phantom data performs well on in vivo data similar to the network fine-tuned on in vivo data. Our results also show that the proposed method outperforms current deep learning methods and is comparable to computationally expensive optimization-based algorithms.

Separated Respiratory Phases for In Vivo Ultrasonic Thermal Strain Imaging

Thermal strain imaging (TSI) uses echo shifts in ultrasonic B-scan images to estimate changes in temperature which is of great values for thermotherapies. However, for in vivo applications, it is difficult to overcome the artifacts and errors arising from physiological motions. Here, a respiration separated TSI (RS-TSI) method is proposed, which can be considered as carrying out TSI in each of the exhalation and inhalation phases and then combining the results. Normalized cross correlation (NXcorr) coefficient between RF images along the timeline are used to extract the respiratory frequency, after which reference frames are selected to identify the exhalation and inhalation phases, and the two phases are divided quasi-periodically. RF images belonging to both phases are selected by applying NXcorr thresholds, and motion compensation together with a second frame selection helps to obtain two finely matched image sequences. After TSI calculations for each phase, the two processes are merged into one through extrapolation and interphase averaging. Compared to TSI based on dynamic frame selection (DFS), RS-TSI ensures that frames are selected during both the exhalation and inhalation phases while setting the frame selection range according to the respiratory frequency helps to improve motion compensation. The temporal intervals of TSI output are approximately half that employing DFS.

External Basic Hyperthermia Devices for Preclinical Studies in Small Animals

Simple Summary

The application of mild hyperthermia can be beneficial for solid tumor treatment by induction of sublethal effects on a tissue- and cellular level. When designing a hyperthermia experiment, several factors should be taken into consideration. In this review, multiple elementary hyperthermia devices are described in detail to aid standardization of treatment design.

Abstract

Preclinical studies have shown that application of mild hyperthermia (40–43 °C) is a promising adjuvant to solid tumor treatment. To improve preclinical testing, enhance reproducibility, and allow comparison of the obtained results, it is crucial to have standardization of the available methods. Reproducibility of methods in and between research groups on the same techniques is crucial to have a better prediction of the clinical outcome and to improve new treatment strategies (for instance with heat-sensitive nanoparticles). Here we provide a preclinically oriented review on the use and applicability of basic hyperthermia systems available for solid tumor thermal treatment in small animals. The complexity of these techniques ranges from a simple, low-cost water bath approach, irradiation with light or lasers, to advanced ultrasound and capacitive heating devices.

Design and Challenges of Sonodynamic Therapy System for Cancer Theranostics: From Equipment to Sensitizers

As a novel noninvasive therapeutic modality combining low-intensity ultrasound and sonosensitizers, sonodynamic therapy (SDT) is promising for clinical translation due to its high tissue-penetrating capability to treat deeper lesions intractable by photodynamic therapy (PDT), which suffers from the major limitation of low tissue penetration depth of light. The effectiveness and feasibility of SDT are regarded to rely on not only the development of stable and flexible SDT apparatus, but also the screening of sonosensitizers with good specificity and safety. To give an outlook of the development of SDT equipment, the key technologies are discussed according to five aspects including ultrasonic dose settings, sonosensitizer screening, tumor positioning, temperature monitoring, and reactive oxygen species (ROS) detection. In addition, some state-of-the-art SDT multifunctional equipment integrating diagnosis and treatment for accurate SDT are introduced. Further, an overview of the development of sonosensitizers is provided from small molecular sensitizers to nano/microenhanced sensitizers. Several types of nanomaterial-augmented SDT are in discussion, including porphyrin-based nanomaterials, porphyrin-like nanomaterials, inorganic nanomaterials, and organic-inorganic hybrid nanomaterials with different strategies to improve SDT therapeutic efficacy. There is no doubt that the rapid development and clinical translation of sonodynamic therapy will be promoted by advanced equipment, smart nanomaterial-based sonosensitizer, and multidisciplinary collaboration.

An Optimized Control Approach for HIFU Tissue Ablation Using PDE Constrained Optimization Method

High-intensity focused ultrasound (HIFU) is a widely used technique capable of providing noninvasive heating and ablation for a wide range of applications. However, the major challenges lie in the determination of the position and the amount of heat deposition over a target area. In order to assure that the thermal area is confined to tumor locations, an optimization method should be employed. Sequential quadratic programming and steepest gradient method with closed-form solution have been previously used to solve this kind of problem. However, these methods are complex and computationally inefficient. The goal of this article is to solve and control the solution of inverse problems with partial differential equation (PDE) constraints. Therefore, a distinguishing challenge of this technique is the handling of large numbers of optimization variables in combination with the complexities of discretized PDEs. In our method, the objective function is formulated as the square difference between the actual thermal dose and the desired one. At each iteration of the optimization procedure, we need to develop and solve the variation problem, the adjoint problem, and the gradient of the objective function. The analytical formula for the gradient is derived and calculated based on the solution of the adjoint problem. Several factors have been taken into consideration to demonstrate the robustness and efficiency of the proposed algorithm. The simulation results for all cases indicate the robustness and the computational efficiency of our proposed method compared to the steepest gradient descent method with the closed-form solution.

Displacement Estimation in Ultrasound Elastography Using Pyramidal Convolutional Neural Network

In this article, two novel deep learning methods are proposed for displacement estimation in ultrasound elastography (USE). Although convolutional neural networks (CNNs) have been very successful for displacement estimation in computer vision, they have been rarely used for USE. One of the main limitations is that the radio frequency (RF) ultrasound data, which is crucial for precise displacement estimation, has vastly different frequency characteristics compared with images in computer vision. Top-rank CNN methods used in computer vision applications are mostly based on a multilevel strategy, which estimates finer resolution based on coarser ones. This strategy does not work well for RF data due to its large high-frequency content. To mitigate the problem, we propose modified pyramid warping and cost volume network (MPWC-Net) and RFMPWC-Net, both based on PWC-Net, to exploit information in RF data by employing two different strategies. We obtained promising results using networks trained only on computer vision images. In the next step, we constructed a large ultrasound simulation database and proposed a new loss function to fine-tune the network to improve its performance. The proposed networks and well-known optical flow networks as well as state-of-the-art elastography methods are evaluated using simulation, phantom, and in vivo data. Our two proposed networks substantially outperform current deep learning methods in terms of contrast-to-noise ratio (CNR) and strain ratio (SR). Also, the proposed methods perform similar to the state-of-the-art elastography methods in terms of CNR and have better SR by substantially reducing the underestimation bias.

Real-time and High Quality Ultrasound Elastography Using Convolutional Neural Network by Incorporating Analytic Signal

Convolutional Neural Networks (CNN) have been extensively used for many computer vision applications including optical flow estimation. Although CNNs have been very successful in optical flow problem, they have been rarely used for displacement estimation in Ultrasound Elastography (USE) due to vast differences between ultrasound data and computer vision images. In USE, a main goal is to obtain the strain image which is the derivative of the axial displacement in axial direction; therefore, a very accurate displacement estimation is required. Radio Frequency (RF) data is needed to obtain accurate displacement estimation. RF data contains high frequency contents which cannot be downsampled without significant loss of information, in contrast to computer vision images. We propose a novel technique to utilize LiteFlowNet for USE. For the first time, we incorporate analytic signal to improve the quality of the displacement estimation. We show that this network with the designed inputs is more suitable for USE compared to more complex networks such as FlowNet2. The network is adopted to our application and it is compared with FlowNet2 and a state-of-the-art elastography method (GLUE). The results show that this network performs well and comparable to GLUE. Furthermore, not only this network is faster and has lower memory footprint compared to FlowNet2, but also it obtains higher quality strain images which makes it suitable for portable and real-time elastography devices.