Deep learning in photoacoustic tomography: current approaches and future directions

Abstract

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.

Oxygen Saturation Imaging Using LED-Based Photoacoustic System

“…Improving the LED array packaging, pulse repetition rate, and coded-excitation schemes may be some aspects to be considered in this direction [ 45 ]. Further, novel deep learning and image processing methods are also expected to have an impact in improving spatial resolution and imaging depth of LED-PAI [ 46 , 47 , 48 ]. In this work, we used a combination LED array of 750 nm and 850 nm with a pulse energy of around 200 J and 100 J per pulse, respectively.…”

Sounding out the hidden data: A concise review of deep learning in photoacoustic imaging

“…Both the fields of photoacoustic imaging and deep learning have progressed at an exceedingly accelerated rate over recent years. [75][76][77] This innovative intersection of fields has served as the staging ground for a number of important innovations in both PACT and PAM, augmenting the capabilities of both imaging methodologies and overcoming many of the persistent challenges facing the field of photoacoustic imaging. We hope this concise review can succinctly summarize recent exciting technological advances and make them accessible to the broader scientific community.…”

Deep Learning in Biomedical Optics

“…Photoacoustic imaging relies on optical transmission, followed by sensing of the resulting acoustic response [194,195]. This response may then be used to guide surgeries and interventions [196,197] (among other possible uses [198]). In order to guide these surgeries and interventions, image maps corresponding to structures of high optical absorption must be formed, which is a rapidly increasing area of interest for the application of DL to photoacoustic imaging and sensing.…”