Limited-view photoacoustic imaging based on an iterative adaptive weighted filtered backprojection approach

Compressed Sensing for Biomedical Photoacoustic Imaging: A Review

Photoacoustic imaging (PAI) is a rapidly developing emerging non-invasive biomedical imaging technique that combines the strong contrast from optical absorption imaging and the high resolution from acoustic imaging. Abnormal biological tissues (such as tumors and inflammation) generate different levels of thermal expansion after absorbing optical energy, producing distinct acoustic signals from normal tissues. This technique can detect small tissue lesions in biological tissues and has demonstrated significant potential for applications in tumor research, melanoma detection, and cardiovascular disease diagnosis. During the process of collecting photoacoustic signals in a PAI system, various factors can influence the signals, such as absorption, scattering, and attenuation in biological tissues. A single ultrasound transducer cannot provide sufficient information to reconstruct high-precision photoacoustic images. To obtain more accurate and clear image reconstruction results, PAI systems typically use a large number of ultrasound transducers to collect multi-channel signals from different angles and positions, thereby acquiring more information about the photoacoustic signals. Therefore, to reconstruct high-quality photoacoustic images, PAI systems require a significant number of measurement signals, which can result in substantial hardware and time costs. Compressed sensing is an algorithm that breaks through the Nyquist sampling theorem and can reconstruct the original signal with a small number of measurement signals. PAI based on compressed sensing has made breakthroughs over the past decade, enabling the reconstruction of low artifacts and high-quality images with a small number of photoacoustic measurement signals, improving time efficiency, and reducing hardware costs. This article provides a detailed introduction to PAI based on compressed sensing, such as the physical transmission model-based compressed sensing method, two-stage reconstruction-based compressed sensing method, and single-pixel camera-based compressed sensing method. Challenges and future perspectives of compressed sensing-based PAI are also discussed.

Adaptive machine learning method for photoacoustic computed tomography based on sparse array sensor data

BACKGROUND AND OBJECTIVE

Photoacoustic computed tomography (PACT) is a non-invasive biomedical imaging technology that has developed rapidly in recent decades, especially has shown potential for small animal studies and early diagnosis of human diseases. To obtain high-quality images, the photoacoustic imaging system needs a high-element-density detector array. However, in practical applications, due to the cost limitation, manufacturing technology, and the system requirement in miniaturization and robustness, it is challenging to achieve sufficient elements and high-quality reconstructed images, which may even suffer from artifacts. Different from the latest machine learning methods based on removing distortions and artifacts to recover high-quality images, this paper proposes an adaptive machine learning method to firstly predict and complement the photoacoustic sensor channel data from sparse array sampling and then reconstruct images through conventional reconstruction algorithms.

METHODS

We develop an adaptive machine learning method to predict and complement the photoacoustic sensor channel data. The model consists of XGBoost and a neural network named SS-net. To handle data sets of different sizes and improve the generalization, a tunable parameter is used to control the weights of XGBoost and SS-net outputs.

RESULTS

The proposed method achieved superior performance as demonstrated by simulation, phantom experiments, and in vivo experiment results. Compared with linear interpolation, XGBoost, CAE, and U-net, the simulation results show that the SSIM value is increased by 12.83%, 6.78%, 21.46%, and 12.33%. Moreover, the median R2 is increased by 34.4%, 8.1%, 28.6%, and 84.1% with the in vivo data.

CONCLUSIONS

This model provides a framework to predict the missed photoacoustic sensor data on a sparse ring-shaped array for PACT imaging and has achieved considerable improvements in reconstructing the objects. Compared with linear interpolation and other deep learning methods qualitatively and quantitatively, our proposed methods can effectively suppress artifacts and improve image quality. The advantage of our methods is that there is no need for preparing a large number of images as the training dataset, and the data for training is directly from the sensors. It has the potential to be applied to a wide range of photoacoustic imaging detector arrays for low-cost and user-friendly clinical applications.

Sparse-view reconstruction for photoacoustic tomography combining diffusion model with model-based iteration

As a non-invasive hybrid biomedical imaging technology, photoacoustic tomography combines high contrast of optical imaging and high penetration of acoustic imaging. However, the conventional standard reconstruction under sparse view could result in low-quality image in photoacoustic tomography. Here, a novel model-based sparse reconstruction method for photoacoustic tomography via diffusion model was proposed. A score-based diffusion model is designed for learning the prior information of the data distribution. The learned prior information is utilized as a constraint for the data consistency term of an optimization problem based on the least-square method in the model-based iterative reconstruction, aiming to achieve the optimal solution. Blood vessels simulation data and the animal in vivo experimental data were used to evaluate the performance of the proposed method. The results demonstrate that the proposed method achieves higher-quality sparse reconstruction compared with conventional reconstruction methods and U-Net. In particular, under the extreme sparse projection (e.g., 32 projections), the proposed method achieves an improvement of ∼ 260 % in structural similarity and ∼ 30 % in peak signal-to-noise ratio for in vivo data, compared with the conventional delay-and-sum method. This method has the potential to reduce the acquisition time and cost of photoacoustic tomography, which will further expand the application range.

Invariant representations in deep learning for optoacoustic imaging

Image reconstruction in optoacoustic tomography (OAT) is a trending learning task highly dependent on measured physical magnitudes present at sensing time. A large number of different settings and also the presence of uncertainties or partial knowledge of parameters can lead to reconstruction algorithms that are specifically tailored and designed to a particular configuration, which could not be the one that will ultimately be faced in a final practical situation. Being able to learn reconstruction algorithms that are robust to different environments (e.g., the different OAT image reconstruction settings) or invariant to such environments is highly valuable because it allows us to focus on what truly matters for the application at hand and discard what are considered spurious features. In this work, we explore the use of deep learning algorithms based on learning invariant and robust representations for the OAT inverse problem. In particular, we consider the application of the ANDMask scheme due to its easy adaptation to the OAT problem. Numerical experiments are conducted showing that when out-of-distribution generalization (against variations in parameters such as the location of the sensors) is imposed, there is no degradation of the performance and, in some cases, it is even possible to achieve improvements with respect to standard deep learning approaches where invariance robustness is not explicitly considered.

Compressed Sensing Photoacoustic Imaging Reconstruction Using Elastic Net Approach

Photoacoustic imaging involves reconstructing an estimation of the absorbed energy density distribution from measured ultrasound data. The reconstruction task based on incomplete and noisy experimental data is usually an ill-posed problem that requires regularization to obtain meaningful solutions. The purpose of the work is to propose an elastic network (EN) model to improve the quality of reconstructed photoacoustic images. To evaluate the performance of the proposed method, a series of numerical simulations and tissue-mimicking phantom experiments are performed. The experiment results indicate that, compared with the L ₁-norm and L ₂-normbased regularization methods with different numerical phantoms, Gaussian noise of 10-50 dB, and different regularization parameters, the EN method with α = 0.5 has better image quality, calculation speed, and antinoise ability.

Approaching closed spherical, full-view detection for photoacoustic tomography

SIGNIFICANCE

Photoacoustic tomography (PAT) is a widely explored imaging modality and has excellent potential for clinical applications. On the acoustic detection side, limited-view angle and limited-bandwidth are common key issues in PAT systems that result in unwanted artifacts. While analytical and simulation studies of limited-view artifacts are fairly extensive, experimental setups capable of comparing limited-view to an ideal full-view case are lacking.

AIMS

A custom ring-shaped detector array was assembled and mounted to a 6-axis robot, then rotated and translated to achieve up to 3.8π steradian view angle coverage of an imaged object.

APPROACH

Minimization of negativity artifacts and phantom imaging were used to optimize the system, followed by demonstrative imaging of a star contrast phantom, a synthetic breast tumor specimen phantom, and a vascular phantom.

RESULTS

Optimization of the angular/rotation scans found ≈212 effective detectors were needed for high-quality images, while 15-mm steps were used to increase the field of view as required depending on the size of the imaged object. Example phantoms were clearly imaged with all discerning features visible and minimal artifacts.

CONCLUSIONS

A near full-view closed spherical system has been developed, paving the way for future work demonstrating experimentally the significant advantages of using a full-view PAT setup.

Artifact removal in photoacoustic tomography with an unsupervised method

Photoacoustic tomography (PAT) is an emerging biomedical imaging technology that can realize high contrast imaging with a penetration depth of the acoustic. Recently, deep learning (DL) methods have also been successfully applied to PAT for improving the image reconstruction quality. However, the current DL-based PAT methods are implemented by the supervised learning strategy, and the imaging performance is dependent on the available ground-truth data. To overcome the limitation, this work introduces a new image domain transformation method based on cyclic generative adversarial network (CycleGAN), termed as PA-GAN, which is used to remove artifacts in PAT images caused by the use of the limited-view measurement data in an unsupervised learning way. A series of data from phantom and in vivo experiments are used to evaluate the performance of the proposed PA-GAN. The experimental results show that PA-GAN provides a good performance in removing artifacts existing in photoacoustic tomographic images. In particular, when dealing with extremely sparse measurement data (e.g., 8 projections in circle phantom experiments), higher imaging performance is achieved by the proposed unsupervised PA-GAN, with an improvement of ∼14% in structural similarity (SSIM) and ∼66% in peak signal to noise ratio (PSNR), compared with the supervised-learning U-Net method. With an increasing number of projections (e.g., 128 projections), U-Net, especially FD U-Net, shows a slight improvement in artifact removal capability, in terms of SSIM and PSNR. Furthermore, the computational time obtained by PA-GAN and U-Net is similar (∼60 ms/frame), once the network is trained. More importantly, PA-GAN is more flexible than U-Net that allows the model to be effectively trained with unpaired data. As a result, PA-GAN makes it possible to implement PAT with higher flexibility without compromising imaging performance.

A Photoacoustic Imaging Algorithm Based on Regularized Smoothed L0 Norm Minimization

The recently emerging technique of sparse reconstruction has received much attention in the field of photoacoustic imaging (PAI). Compressed sensing (CS) has large potential in efficiently reconstructing high-quality PAI images with sparse sampling signal. In this article, we propose a CS-based error-tolerant regularized smooth L0 (ReSL0) algorithm for PAI image reconstruction, which has the same computational advantages as the SL0 algorithm while having a higher degree of immunity to inaccuracy caused by noise. In order to evaluate the performance of the ReSL0 algorithm, we reconstruct the simulated dataset obtained from three phantoms. In addition, a real experimental dataset from agar phantom is also used to verify the effectiveness of the ReSL0 algorithm. Compared to three L0 norm, L1 norm, and TV norm-based CS algorithms for signal recovery and image reconstruction, experiments demonstrated that the ReSL0 algorithm provides a good balance between the quality and efficiency of reconstructions. Furthermore, the PSNR of the reconstructed image calculated by the introduced method was better than the other three methods. In particular, it can notably improve reconstruction quality in the case of noisy measurement.

LV-GAN: A deep learning approach for limited-view optoacoustic imaging based on hybrid datasets

The optoacoustic imaging (OAI) methods are rapidly evolving for resolving optical contrast in medical imaging applications. In practice, measurement strategies are commonly implemented under limited-view conditions due to oversized image objectives or system design limitations. Data acquired by limited-view detection may impart artifacts and distortions in reconstructed optoacoustic (OA) images. We propose a hybrid data-driven deep learning approach based on generative adversarial network (GAN), termed as LV-GAN, to efficiently recover high quality images from limited-view OA images. Trained on both simulation and experiment data, LV-GAN is found capable of achieving high recovery accuracy even under limited detection angles less than 60^° . The feasibility of LV-GAN for artifact removal in biological applications was validated by ex vivo experiments based on two different OAI systems, suggesting high potential of a ubiquitous use of LV-GAN to optimize image quality or system design for different scanners and application scenarios.

A generative adversarial network for artifact removal in photoacoustic computed tomography with a linear-array transducer

With balanced spatial resolution, penetration depth, and imaging speed, photoacoustic computed tomography (PACT) is promising for clinical translation such as in breast cancer screening, functional brain imaging, and surgical guidance. Typically using a linear ultrasound (US) transducer array, PACT has great flexibility for hand-held applications. However, the linear US transducer array has a limited detection angle range and frequency bandwidth, resulting in limited-view and limited-bandwidth artifacts in the reconstructed PACT images. These artifacts significantly reduce the imaging quality. To address these issues, existing solutions often have to pay the price of system complexity, cost, and/or imaging speed. Here, we propose a deep-learning-based method that explores the Wasserstein generative adversarial network with gradient penalty (WGAN-GP) to reduce the limited-view and limited-bandwidth artifacts in PACT. Compared with existing reconstruction and convolutional neural network approach, our model has shown improvement in imaging quality and resolution. Our results on simulation, phantom, and in vivo data have collectively demonstrated the feasibility of applying WGAN-GP to improve PACT’s image quality without any modification to the current imaging set-up.

Impact statement

This study has the following main impacts. It offers a promising solution for removing limited-view and limited-bandwidth artifact in PACT using a linear-array transducer and conventional image reconstruction, which have long hindered its clinical translation. Our solution shows unprecedented artifact removal ability for in vivo image, which may enable important applications such as imaging tumor angiogenesis and hypoxia. The study reports, for the first time, the use of an advanced deep-learning model based on stabilized generative adversarial network. Our results have demonstrated its superiority over other state-of-the-art deep-learning methods.

Image reconstruction based on compressed sensing for sparse-data endoscopic photoacoustic tomography

Endoscopic photoacoustic tomography (EPAT) is an interventional application of photoacoustic tomography (PAT) to visualize anatomical features and functional components of biological cavity structures such as nasal cavity, digestive tract or coronary arterial vessels. One of the main challenges in clinical applicability of EPAT is the incomplete acoustic measurements due to the limited detectors or the limited-view acoustic detection enclosed in the cavity. In this case, conventional image reconstruction methodologies suffer from significantly degraded image quality. This work introduces a compressed-sensing (CS)-based method to reconstruct a high-quality image that represents the initial pressure distribution on a luminal cross-section from incomplete discrete acoustic measurements. The method constructs and trains a complete dictionary for the sparse representation of the photoacoustically-induced acoustic measurements. The sparse representation of the complete acoustic signals is then optimally obtained based on the sparse measurements and a sensing matrix. The complete acoustic signals are recovered from the sparse representation by inverse sparse transformation. The image of the initial pressure distribution is finally reconstructed from the recovered complete signals by using the time reversal (TR) algorithm. It was shown with numerical experiments that high-quality images with reduced under-sampling artifacts can be reconstructed from sparse measurements. The comparison results suggest that the proposed method outperforms the standard TR reconstruction by 40% in terms of the structural similarity of the reconstructed images.

Wave front analysis for enhanced time-domain beamforming of point-like targets in optoacoustic imaging using a linear array

Using linear array transducers in combination with state-of-the-art multichannel electronics allows to perform optoacoustic imaging with frame rates only limited by the laser pulse repetition frequency and the acoustic time of flight. However, characteristic image artefacts resulting from the limited view and a lower SNR when compared to systems based on single-element focused transducers represent a burden for the clinical acceptance of the technology. In this paper, we present a new method for the improvement of image quality based on the analysis of the signal amplitudes along summation curves during the delay-and-sum beamforming process (DAS). The algorithm compares amplitude distributions along wave fronts with theoretical patterns from optoacoustic point sources. The method was validated on simulated and experimental phantom as well as in-vivo data. An improvement of the lateral resolution by more than a factor of two when comparing conventional DAS and our approach could be shown (numeric and experimental phantom data). For instance, on experimental data from a wire phantom, a PSF in the range of 0.18-0.22 mm was obtained with our approach against 0.48 mm for standard DAS. Furthermore, the SNR of a subcutaneous vessel 2.5 mm below the skin surface was improved by about 30 dB when compared to standard DAS.

Improved back-projection method for circular-scanning-based photoacoustic tomography with improved tangential resolution

While photoacoustic computed tomography (PACT) is generally built with planar transducers of finite size, most current reconstruction algorithms assume the transducer to be an ideal point, which leads to a spinning blur in the consequently obtained PACT images due to the model mismatch. In this work, we put forward an improved back-projection method that factors in the geometry of the transducers to improve the tangential resolution for the reconstruction of 2D circular-scanning-based photoacoustic tomography. Extensive simulations and experiments were carried out to study the adaptability and stability of the proposed method. Results show that this method can effectively restore the tangential distortion of the PACT image for both simulated and experimental data. Results indicated that the improvement of the tangential resolution is more obvious for transducers with larger size. We also demonstrated the application of this method to transducers other than planar, and results show that the reconstructed image quality can be significantly affected by the shape and position of the transducers used. This study may help to guide the selection of transducer and design of scanning strategy in PACT.

On the link between the speckle free nature of optoacoustics and visibility of structures in limited-view tomography

Similar to pulse-echo ultrasound, optoacoustic imaging encodes the location of optical absorbers by the time-of-flight of ultrasound waves. Yet, signal generation mechanisms are fundamentally different for the two modalities, leading to significant distinction between the optimum image formation strategies. While interference of back-scattered ultrasound waves with random phases causes speckle noise in ultrasound images, speckle formation is hindered by the strong correlation between the optoacoustic responses corresponding to individual sources. However, visibility of structures is severely hampered when attempting to acquire optoacoustic images under limited-view tomographic geometries. In this tutorial article, we systematically describe the basic principles of optoacoustic signal generation and image formation for objects ranging from individual sub-resolution absorbers to a continuous absorption distribution. The results are of relevance for the proper interpretation of optoacoustic images acquired under limited-view scenarios and may also serve as a basis for optimal design of tomographic acquisition geometries and image formation strategies.

Regularized Iterative Weighted Filtered Back-Projection for Few-View Data Photoacoustic Imaging

Photoacoustic imaging is an emerging noninvasive imaging technique with great potential for a wide range of biomedical imaging applications. However, with few-view data the filtered back-projection method will create streak artifacts. In this study, the regularized iterative weighted filtered back-projection method was applied to our photoacoustic imaging of the optical absorption in phantom from few-view data. This method is based on iterative application of a nonexact 2DFBP. By adding a regularization operation in the iterative loop, the streak artifacts have been reduced to a great extent and the convergence properties of the iterative scheme have been improved. Results of numerical simulations demonstrated that the proposed method was superior to the iterative FBP method in terms of both accuracy and robustness to noise. The quantitative image evaluation studies have shown that the proposed method outperforms conventional iterative methods.

Ultrasonic-heating-encoded photoacoustic tomography with virtually augmented detection view

Photoacoustic (PA) imaging of arbitrarily-shaped or oriented objects may miss important features because PA waves propagate normal to structure boundaries and may miss the acoustic detectors when the detection view has a limited angular range. To overcome this long-standing problem, we present an ultrasonic thermal encoding approach that is universally applicable. We exploit the temperature dependence of the Grueneisen parameter and encode a confined [[What does confined mean here?]] voxel using heat generated by a focused ultrasonic transducer. The PA amplitude from the encoded voxel is increased while those from the neighboring voxels are unchanged. Consequently, the amplitude-increased PA waves propagate in all directions due to the round cross-section of the encoded region and thus can be received at any viewing angle on the cross-sectional plane [[Please check throughout the manuscript for similar places.]]. We built a mathematical model for the thermally encoded PA tomography, performed a numerical simulation, and experimentally validated the ultrasonic thermal encoding efficiency. As a proof of concept, we demonstrate full-view in vivo vascular imaging and compare it to the original linear-array PA tomography system, showing dramatically enhanced imaging of arbitrarily oriented blood vessels. Since ultrasonic heating can be focused deeply, this method can be applied to deep tissue imaging and is promising for full-view imaging of other features of biomedical interest, such as tumor margins.