Artifacts in Photoacoustic Imaging: Origins and Mitigations

Max T Rietberg, Janek Gröhl, Thomas R Else, Sarah E Bohndiek, Srirang Manohar, Benjamin T Cox · 2025

Read Artifacts in Photoacoustic Imaging: Origins and Mitigations by Max T Rietberg, Janek Gröhl, Thomas R Else, Sarah E Bohndiek, Srirang Manohar, Benjamin T Cox (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-12772.txt.

Introduction

Max Rietberg and co-authors catalog the artifacts that mar photoacoustic images in a 2025 review. The technique shines light into tissue and listens to the faint sound the heated spots emit, building pictures of blood and pigment. Devices are now approved for clinic use, so flaws that could mislead care need attention.

The authors explain the physics, then sort artifacts by cause, with sketches and both simulated and real examples. Two roots dominate: systems cannot record the full data set, and rebuilding the image leans on simplifying assumptions. The paper serves two audiences at once, teaching clinicians to spot and judge artifacts, and giving engineers a reference of current limits. It closes with advanced ways to reduce the flaws.

Catalog record

CREATORMax T Rietberg, Janek Gröhl, Thomas R Else, Sarah E Bohndiek, Srirang Manohar, Benjamin T Cox
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE225 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-12772

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Artifacts in Photoacoustic Imaging: Origins and Mitigations Max T Rietberg, Janek Gröhl, Thomas R Else, Sarah E Bohndiek, Srirang Manohar, Benjamin T Cox Abstract Photoacoustic imaging (PAI) is rapidly moving from the laboratory to the clinic, increasing the need to understand confounders which might adversely affect patient care. Over the past five years, landmark studies have shown the clinical utility of PAI, leading to regulatory approval of several devices. In this article, we describe the various causes of artifacts in PAI, providing schematic overviews and practical examples, simulated as well as experimental. This work serves two purposes: (1) educating clinical users to identify artifacts, understand their causes, and assess whether their impact, and (2) providing a reference of the limitations of current systems for those working to improve them. We explain how two aspects of PAI systems lead to artifacts: their inability to measure complete data sets, and embedded assumptions during reconstruction. We describe the physics underlying PAI, and propose a classification of the artifacts. The paper concludes by discussing possible advanced mitigation strategies. Journal: arXiv 1 Introduction Medical imaging technologies enable visualization of the structure and function of biological tissues, thereby facilitating disease detection, diagnosis, monitoring of treatment response and follow-up. Medical imaging modalities rely on the measurement of transmitted, absorbed, scattered, or emitted energy of some form. Measured data and physical models of energy-tissue interactions are then used to reconstruct images [ 1 ] . These models necessarily involve simplifying assumptions, and so describe the true physical processes only approximately. Furthermore, during data collection some information will be lost, whether due to hardware limitations or subject and environmental factors. The reconstructed image will therefore not be a perfect representation of the underlying tissue but will contain deviations, called artifacts. An artifact is a structure, distortion, or feature in an image that does not correspond to the actual anatomical, morphological, physiological, or pathological reality of the imaged subject [ 2 , 3 ] . The presence of artifacts may confound clinical interpretation and negatively affect decision making. For example, artifacts includes the obscuring, displacement or distortion of genuine structural or functional features, or the presence of features that are not real, perhaps giving the illusion of pathophysiology. Artifacts can lead to images being misinterpreted, errors in quantitative (e.g. molecular) information extracted from the image, and false positive or false negative imaging biomarkers. A thorough understanding of artifacts in medical imaging is therefore crucial to ensure reliable and accurate clinical use. Artifacts can be especially prominent in multi-physics imaging modalities like photoacoustic imaging. In photoacoustic (PA) imaging (PAI) the acoustic emissions from the absorbed optical energy distribution are measured, in contrast to purely optical imaging methods such as diffuse optical tomography [ 4 ] , where transmitted or reflected light is detected. In PAI, tissue is interrogated with nanosecond pulsed light, which is scattered and absorbed inside the tissue. The light is absorbed by chromophores, leading to a localized temperature and pressure rise, and subsequently, due to thermoelastic expansion, the emergence of acoustic pressure waves [ 5 ] . Detection of acoustic waves outside the tissue followed by image reconstruction enables visualization of the distribution of absorbed optical energy up to several centimeters deep. By varying the wavelength of incident light, absorption of several different molecules can be probed [ 6 ] . Typical features extracted from PAI with promise for clinical translation include (1) vascular architecture [ 7 , 8 , 9 ] , (2) signal intensity at a target wavelength [ 10 , 11 , 12 ] , (3) blood oxygen saturation [ 13 , 14 , 15 ] , (4) general molecular concentration information of, e.g., fatty tissue, glandular tissue, water [ 16 , 17 ] , and (5) contrast agent distribution [ 18 , 19 ] . Photoacoustic image formation thus involves an interplay of optical, thermal, elastic and acoustic properties, and gives rise to a variety of unique artifacts not seen in optical or ultrasound imaging separately. The presence of artifacts can have negative effects on the qualitative and quantitative interpretation of PAI images and their extracted features. In the context of PAI, we define an artifact as a feature in the image that does not correspond, spatially or spectrally, to the true anatomical, morphological and spectral characteristics of the imaging target under investigation. As PAI is on the cusp of achieving clinical breakthroughs in various applications [ 20 , 21 ] , we believe it would be helpful to raise awareness of artifacts in the PAI community, especially among the growing number of clinical PAI users. PA users should be able to identify artifacts, so they can better understand and improve the interpretation of in vivo PA images or even take advantage of the artifacts, as they could carry diagnostic information. Furthermore, we hope that recognizing artifacts and understanding their origins will help guide the development of PAI systems towards mitigations and hasten the progress towards improved PAI technology. In this work, we structure the manuscript by the sources and causes of artifacts. Starting with a short description of the physics involved in PAI, we continue by outlining the reasons behind the emergence of each artifact, and briefly discuss possible mitigation strategies. Each artifact is demonstrated with simulated images and, where available, experimental examples. Simulations are performed on simple digital phantoms in which the assumptions made during image reconstruction are deliberately broken to demonstrate their confounding effect. For illustrative purposes, the simulations are restricted to simple but typical clinical PA detection geometries; however, similar artifacts would appear in other hardware configurations [ 22 , 23 ] . We conclude with a discussion of possible advanced artifact mitigation strategies. 2 Artifact Classification Necessary compromises in the design of PAI systems mean that most PA images contain artifacts. Taking a high-level view, there are two parts to any PAI system: hardware for data acquisition, and a method for reconstructing and displaying the images. There are two corresponding classes of underlying causes of artifacts: (1) insufficient data, e.g. limited array size or limited bandwidth, and (2) incorrect assumptions in the image reconstruction algorithm, e.g. inaccurate approximations to the physics of wave propagation or an overly simple model of the detector response. In practical applications, there are often good reasons for designing an imaging system that measures incomplete data, or for developing an image reconstruction approach that makes inaccurate assumptions. First, there are clinical requirements . These are dictated by what the clinician or researcher needs to see, which requires the instrument to access the body part of interest with a sufficient field-of-view and resolution. Further challenges, such as restrictions on the clinical workspace, fitting around other devices, and patient safety and comfort, are also key considerations. Second, there are physics constraints , e.g. limited light penetration or acoustic attenuation that limit imaging depths. Third, there are practical limitations , which include the availability of resources (raw materials, fabrication facilities, computing hardware), financial costs of hardware and operation, or even intellectual property rights and market demand. Trade-offs must be made in imaging system design, leading to widely differing PAI systems from microscopes to large-area bowls for breast imaging. Artifacts in PAI may be classified in many different ways.

Read the complete text