Artifact removal in photoacoustic section imaging by combining an integrating cylindrical detector with model-based reconstruction

Guenther Paltauf,Robert Nuster

doi:10.1117/1.jbo.19.2.026014

Abstract

Photoacoustic section imaging reveals optically absorbing structures within a thin slice of an object. It requires measuring acoustic waves excited by absorption of short laser pulses with a cylindrical acoustic lens detector rotating around the object. Owing to the finite detector size and its limited depth of focus, various artifacts arise, seen as distortions within the imaging slice and cross-talk from neighboring areas of the object. The presented solution aims at avoiding these artifacts by a special design of the sensor and by use of a model-based reconstruction algorithm that improves section images by incorporating information from neighboring sections. The integrating property of the cylindrical detector, which exceeds in direction of the cylinder axis the size of the imaged object, avoids the lateral blurring that normally results from the finite width of a small detector. Applying a maximum likelihood reconstruction method for the inversion of the imaging system matrix to the temporal pressure signals yields line projections of the initial energy distribution, from which section images are obtained by applying the inverse Radon transform. By using data from few sections, a significant reduction of artifacts related to the imperfections of the sensor is demonstrated both in simulations and in phantom experiments.

Highlights

Photoacoustic [(PA) or optoacoustic] imaging is a technique for visualization of structures with optical absorption contrast in light scattering media
A horizontal profile through one of the sources is displayed together with the reconstruction that is obtained by applying the inverse Radon transform to Abel-transformed pressure signals
The residual peak after lsqr reconstruction reaches a value of ∼0.1, whereas the maximum likelihood-expectation maximization (ML-EM) reconstruction yields a negligible value on the order of 0.05

Summary

Introduction

Photoacoustic [(PA) or optoacoustic] imaging is a technique for visualization of structures with optical absorption contrast in light scattering media. It is based on the excitation of acoustic waves by the thermoelastic effect, when nanosecond duration electromagnetic pulses are absorbed in an object. The distribution of absorbed energy is obtained from broad band ultrasound signals measured outside the object, from which structures with preferential light absorption can be localized. Because of the ability to visualize the interior of optically scattering objects with ultrasound resolution and the good contrast of blood vessels, PA methods have been recognized as a promising imaging tool for biological tissue.[1]. The distribution of absorbed energy density is obtained by applying a tomographic reconstruction algorithm to the signals. No tomographic reconstruction is required as the single amplitude scans can be combined to an image plane or volume for two-dimensional (2-D) and three-dimensional (3-D) imaging, respectively

Objectives

Methods

Results

Discussion

Conclusion