Model-Based Learning for Accelerated, Limited-View 3-D Photoacoustic Tomography

Abstract

Recent advances in deep learning for tomographic reconstructions have shown great potential to create accurate and high quality images with a considerable speed up. In this paper, we present a deep neural network that is specifically designed to provide high resolution 3-D images from restricted photoacoustic measurements. The network is designed to represent an iterative scheme and incorporates gradient information of the data fit to compensate for limited view artifacts. Due to the high complexity of the photoacoustic forward operator, we separate training and computation of the gradient information. A suitable prior for the desired image structures is learned as part of the training. The resulting network is trained and tested on a set of segmented vessels from lung computed tomography scans and then applied to in-vivo photoacoustic measurement data.

Multi-Scale Learned Iterative Reconstruction

“…On the downside, large capacity networks tend to overfit to the training data and especially so when the training data is scarce. Furthermore, as shown in [12], [17], [18], [30] the results are clearly outperformed by learned iterative reconstruction algorithms that we next describe.…”

“…In gradient boosting, that follow the greedy training [18], the loss function is changed. Instead of looking for a reconstruction operator that is optimal end-to-end, we only require iterate-wise optimality.…”

“…One possible solution to address these challenges is to adopt a greedy approach for training. Here each unrolled iteration in the network is trained separately [18]. In this way, training of each unrolled iterate and evaluation of the forward operator can be separated, thus rendering a training procedure feasible.…”

Learned backprojection for sparse and limited view photoacoustic tomography

“…Deep learning and in particular deep convolutional neural networks (CNNs) have been very successfully applied to a great variety of pattern recognition and image processing tasks. Recently a lot of research has been done in solving inverse problems, incorporating deep learning techniques, including efficient and accurate image reconstruction methods in tomographic problems [2,5,12,13,14,28,17].…”

“…It is known, that the reconstruction of singularities not visible from the detector positions is unstable, whereas singularities contained within the convex hull can theoretically be stably recovered in appropriate spaces [15]. For PAT from limited view data several iterative reconstruction methods have been proposed to reduce limited view artifacts and improve the reconstruction quality [22,7,13]. In [21] the authors proposed using weight factors depending on the angle between the reconstruction-and detection point for an inversion formula, which is exact on continuously sampled data given on the whole boundary.…”