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

Abstract

The rapidly evolving field of photoacoustic tomography utilizes endogenous chromophores to extract both functional and structural information from deep within tissues. It is this power to perform precise quantitative measurements in vivo—with endogenous or exogenous contrast —that makes photoacoustic tomography highly promising for clinical translation in functional brain imaging, early cancer detection, real-time surgical guidance, and the visualization of dynamic drug responses. Considering photoacoustic tomography has benefited from numerous engineering innovations, it is of no surprise that many of photoacoustic tomography’s current cutting-edge developments incorporate advances from the equally novel field of artificial intelligence. More specifically, alongside the growth and prevalence of graphical processing unit capabilities within recent years has emerged an offshoot of artificial intelligence known as deep learning. Rooted in the solid foundation of signal processing, deep learning typically utilizes a method of optimization known as gradient descent to minimize a loss function and update model parameters. There are already a number of innovative efforts in photoacoustic tomography utilizing deep learning techniques for a variety of purposes, including resolution enhancement, reconstruction artifact removal, undersampling correction, and improved quantification. Most of these efforts have proven to be highly promising in addressing long-standing technical obstacles where traditional solutions either completely fail or make only incremental progress. This concise review focuses on the history of applied artificial intelligence in photoacoustic tomography, presents recent advances at this multifaceted intersection of fields, and outlines the most exciting advances that will likely propagate into promising future innovations.

Sound out the impaired perfusion: Photoacoustic imaging in preclinical ischemic stroke

“…PAM technology has seen improvement in imaging speed, field-of-view, resolution, and sensitivity ( Wang et al, 2021 ). These improvements have come from the application of novel high-speed multi-spectra light sources ( Cho et al, 2018 , 2021 ; Chen et al, 2022 ), deep learning imaging enhancement ( DiSpirito et al, 2020 , DiSpirito et al, 2021 ), innovative scanning configurations ( Liu and Yao, 2018 ), and exogenous contrast agents ( Borg and Rochford, 2018 ). These developments have allowed for better quantification and visualization of the hemodynamics of ischemic stroke.…”

Perspective on fast-evolving photoacoustic tomography

“…Limited by space, we will focus on several important developments that have overcome the longstanding limits in traditional PAT. Interested readers are referred to comprehensive review articles that provide in-depth analyses and discussions on low-cost light sources, 22 , 26 – 28 novel ultrasound sensors, 21 , 29 , 30 PA contrast agents, 29 , 31 , 32 PA endoscopy, 33 , 34 deep learning enhanced PAT, 35 – 37 as well as clinical translation of PAT. 22 , 38 – 41 …”

U‐Net enhanced real‐time LED‐based photoacoustic imaging

“…Specifically in the realm of machine learning strategies, many deep-learning network models are being investigated to improve PA image reconstruction methodologies [20][21][22][23]. Several important studies, including recent reviews [24][25][26][27] in deep learning with laser-based systems, addressed the problem of under-sampled data sparsity due to the limited number of detectors. While many studies demonstrated the results on numerically simulated data and in vitro phantoms [28][29][30][31][32][33][34][35], only a handful of studies tested their deep-learning models on in vivo data.…”