Modeling of Errors Due to Uncertainties in Ultrasound Sensor Locations in Photoacoustic Tomography

Biomedical photoacoustic tomography, which can provide high-resolution 3D soft tissue images based on optical absorption, has advanced to the stage at which translation from the laboratory to clinical settings is becoming possible. The need for rapid image formation and the practical restrictions on data acquisition that arise from the constraints of a clinical workflow are presenting new image reconstruction challenges. There are many classical approaches to image reconstruction, but ameliorating the effects of incomplete or imperfect data through the incorporation of accurate priors is challenging and leads to slow algorithms. Recently, the application of deep learning (DL), or deep neural networks, to this problem has received a great deal of attention. We review the literature on learned image reconstruction, summarizing the current trends and explain how these approaches fit within, and to some extent have arisen from, a framework that encompasses classical reconstruction methods. In particular, it shows how these techniques can be understood from a Bayesian perspective, providing useful insights. We also provide a concise tutorial demonstration of three prototypical approaches to learned image reconstruction. The code and data sets for these demonstrations are available to researchers. It is anticipated that it is in in vivo applications-where data may be sparse, fast imaging critical, and priors difficult to construct by hand-that DL will have the most impact. With this in mind, we conclude with some indications of possible future research directions. © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

“…This approach has been applied in PAT to compensate for uncertainty in the measurement parameters (model uncertainty). 41,42…”

Section: Statistical Framework: Noise Models and Priorsmentioning

confidence: 99%

Deep learning in photoacoustic tomography: current approaches and future directions

2020

View full text Add to dashboard Cite

show abstract

“…The assumptions for non-attenuating and homogeneous domain have been shown to produce accurate photoacoustic images in soft-tissue or soft-tissue mimicking targets [14], [15], [17]- [19]. However, if the target is composed of heterogeneous tissues such as soft-tissue and bone, the homogeneous model could still be useful when the modeling errors are taken into account [30].…”

Section: Forward Modelmentioning

confidence: 99%

“…In this work, the inverse problem of PAT is approached in a Bayesian framework [17], [19], [35]. In the approach, all parameters are modeled as random variables, and it combines the information obtained through the measurements, forward model, and prior model for the unknown parameters.…”

Section: Inverse Problemmentioning

confidence: 99%

“…The parameters κ and q control the shape of the covariance function. Thus by the choice of these parameters, the covariance function ( 25) can be tuned to closely match the Ornstein-Uhlenbeck process that has previously been utilized in PAT [17]- [19], [32]. The advantage of the piece-wise polynomial covariance function, compared to the Ornstein-Uhlenbeck process, is that the values of the covariance become exactly zero after some distance κ.…”

Section: A Prior Distributionmentioning

confidence: 99%

See 1 more Smart Citation

Computationally Efficient Forward Operator for Photoacoustic Tomography Based on Coordinate Transformations

Sahlström

Pulkkinen

Leskinen

et al. 2021

IEEE Trans. Ultrason., Ferroelect., Freq. Contr.

Self Cite

View full text Add to dashboard Cite

Photoacoustic tomography (PAT) is an imaging modality that utilizes the photoacoustic effect. In PAT, a photoacoustic image is computed from measured data by modeling ultrasound propagation in the imaged domain and solving an inverse problem utilizing a discrete forward operator. However, in realistic measurement geometries with several ultrasound transducers and relatively large imaging volume, an explicit formation and use of the forward operator can be computationally prohibitively expensive. In this work, we propose a transformation based approach for efficient modeling of photoacoustic signals and reconstruction of photoacoustic images. In the approach, the forward operator is constructed for a reference ultrasound transducer and expanded into a general measurement geometry using transformations that map the formulated forward operator in local coordinates to the global coordinates of the measurement geometry. The inverse problem is solved using a Bayesian framework. The approach is evaluated with numerical simulations and experimental data. The results show that the proposed approach produces accurate three-dimensional photoacoustic images with a significantly reduced computational cost both in memory requirements and in time. In the studied cases, depending on the computational factors such as discretization, over 30-fold reduction in memory consumption and was achieved without a reduction in image quality compared to a conventional approach.

show abstract

“…Compensation of the modelling errors is carried out using Bayesian approximation error modelling. 9,10 The rest of the paper is structured as follows. The forward model for PAT is described in Section 2.…”

Section: Introductionmentioning

confidence: 99%

Modelling of uncertainties in ultrasound sensor locations in photoacoustic tomography

Sahlström

Pulkkinen

Tick

et al. 2020

Photons Plus Ultrasound: Imaging and Sensing 2020

Self Cite

View full text Add to dashboard Cite

Photoacoustic tomography (PAT) is an imaging modality developed during the past few decades. In the inverse problem of PAT, the aim is to estimate the spatial distribution of an initial pressure p 0 generated by the photoacoustic effect, when photoacoustic time-series p t measured on the boundary of the imaged target are given. To produce accurate photoacoustic images, the forward model linking p 0 to p t has to model the measurement setup and the underlying physics to a sufficient accuracy. Use of an inaccurate model can lead to significant errors in the solution of the inverse problem. In this work, we study the effect and compensation of modelling errors due to uncertainties in ultrasound sensor locations in PAT using Bayesian approximation error modelling. The approach is evaluated with simulated and experimental data using various levels of measurement noise, uncertainties in sensor locations and varying sensor geometries. The results indicate that even small errors in the modelling of ultrasound sensor locations can lead to large errors in the solution of the inverse problem. Furthermore, the magnitude of these errors is affected by the amount of measurement noise and the measurement geometry. The modelling errors can, however, be well compensated by the approximation error modelling.

show abstract

Modeling of Errors Due to Uncertainties in Ultrasound Sensor Locations in Photoacoustic Tomography

Cited by 26 publications

References 41 publications

Deep learning in photoacoustic tomography: current approaches and future directions

Deep learning in photoacoustic tomography: current approaches and future directions

Computationally Efficient Forward Operator for Photoacoustic Tomography Based on Coordinate Transformations

Modelling of uncertainties in ultrasound sensor locations in photoacoustic tomography

Contact Info

Product

Resources

About