Proc. SPIE. 10878, Photons Plus Ultrasound: Imaging and Sensing 2019
KEYWORDS: Signal to noise ratio, Real time imaging, Image restoration, Interference (communication), Computer simulations, Transducers, Photoacoustic imaging, Reconstruction algorithms, Photoacoustic spectroscopy, Signal detection
Delay and Sum (DAS) is one of the most common beamforming algorithms for photoacoustic imaging reconstruction that can function well in real-time imaging for its simplicity and quickness. However, high sidelobes and intense artifacts usually appear in the reconstructed image using DAS algorithm. To solve this problem, a novel beamforming algorithm called Multiple Delay and Sum with Enveloping (multi-DASE) is introduced in this paper, which can suppress sidelobes and artifacts efficiently. Compared to DAS, multi-DASE beamforming algorithm calculates not only the initial beamformed signal but also the N-shaped photoacoustic signal for each pixel. Firstly, Delay and Sum is performed multiply based on time series to recover the N-shaped photoacoustic signal for each pixel in the reconstructed image. And then, the recovered signal is enveloped to transform the N-shaped wave into a pulse wave to remove the negative part of the signal. Finally, signal suppression is performed on the enveloped signal which can lead to the suppression of sidelobes and artifacts in the reconstructed image. The multi-DASE beamforming algorithm was tested on the simulated data acquired with MATLAB k-Wave Toolbox. Experiment was also conducted to evaluate the efficiency of the multiDASE algorithm for clinical application. Both in computer simulation and experiment, our multi-DASE beamforming algorithm showed great performance in removing artifacts and improving image quality. In our multi-DASE beamforming algorithm, only fundamental operations and Discrete Fourier Transform (DFT) are performed, which means it can be a promising method for real-time clinical application.
Hyperosteogeny and Osteoporosis are two common bone diseases that have a high incidence in the middle-aged and elderly groups. Mild symptoms may only affect the daily life of patients, while severe ones are life-threatening. At present, detection methods based on X-ray film and ultrasound are generally applied. However, the former exist errors introduced by manual reading and a certain radiation hazard, the diagnostic results of the latter are not that satisfying as well. Photoacoustic effect combines the advantages of optics for sensitive light absorption contrast and acoustics for lower acoustic scattering in soft tissue. As a non-ionizing and non-invasive technique, its application in biomedicine is also emerging. In this paper, a classification model built on Convolutional Neural Network (CNN) was proposed to achieve automated diagnosis of hyperosteogeny, osteoporosis and normal bone. Time-domain photoacoustic signals generated by different bone types are set as the inputs of the CNN while the output results indicate the corresponding categories of the samples. The analysis results of ex vivo data demonstrated that the established model could accurately accomplish the research of classification. Thus, the proposed method has certain auxiliary value for improving the efficiency, accuracy and objectivity of clinical diagnosis of the three bone types.
Photoacoustic (PA) signal of an ideal optical absorb particle is a single N-shape wave. PA signals of a complicated biological tissue can be considered as the combination of individual N-shape waves. However, the N-shape wave basis not only complicates the subsequent work, but also results in aliasing between adjacent micro-structures, which deteriorates the quality of the final PA images. In this paper, we propose a method to improve PA image quality through signal processing method directly working on raw signals, which including deconvolution and empirical mode decomposition (EMD). During the deconvolution procedure, the raw PA signals are de-convolved with a system dependent point spread function (PSF) which is measured in advance. Then, EMD is adopted to adaptively re-shape the PA signals with two constraints, positive polarity and spectrum consistence. With our proposed method, the built PA images can yield more detail structural information. Micro-structures are clearly separated and revealed. To validate the effectiveness of this method, we present numerical simulations and phantom studies consist of a densely distributed point sources model and a blood vessel model. In the future, our study might hold the potential for clinical PA imaging as it can help to distinguish micro-structures from the optimized images and even measure the size of objects from deconvolved signals.