Research On Photoacoustic Tomography Based On Total Variational Regularization

Medical imaging technology is an attractive diagnostic tool that can collect data from biological tissues.Photoacoustic imaging is an emerging medical imaging technology that combines the advantages of optical imaging and ultrasound imaging.It uses laser pulses to excite ultrasound signals to obtain information on the distribution of optical absorption in biological tissues,thereby reconstructing images.Choosing the appropriate image reconstruction algorithm is one of the key steps to achieve high-quality image reconstruction in photoacoustic tomography.This article aims to study image reconstruction algorithms in photoacoustic tomography to find algorithms that can improve image reconstruction quality and speed up image reconstruction under different models.The main content of the paper is as follows:(1)The forward process of photoacoustic imaging was introduced,and the Green’s function method was used to solve the photoacoustic wave equation.In the inverse process of photoacoustic imaging,three types of inversion algorithms were introduced: analytical,time reversal,and iterative algorithms,and the geometric shapes to which each algorithm is applicable were explained.(2)To address issues such as slow image reconstruction speed and low resolution,this article adopts the Delay and Sum(DAS)algorithm,which calculates the light absorption intensity at a given location based on the difference in distance between that location and the position of the detector.The Time Reversal(TR)algorithm is also used,which utilizes the time-reversal invariance of photoacoustic signals propagating through tissues to eliminate artifacts caused by acoustic scattering and improve image resolution.(3)In order to address the issue of edge blurring in photoacoustic tomography image reconstruction,this article proposes a TV-CG algorithm that combines total variation regularization and conjugate gradient method.By introducing TV perturbation in the conjugate gradient iteration process,the performance of the reconstructed image is effectively improved.Compared with DAS and TR algorithms,TV-CG algorithm has similar reconstruction time.Experimental results show that this algorithm is suitable for reconstructing original images with relatively smooth edges.(4)This article introduces a non-convex total variation regularization(TV-L2)model based on regularization for improving the image reconstruction quality in photoacoustic computed tomography.By constructing a non-convex regularization norm and applying it to the total variation regularization,a non-convex total variation regularization reconstruction model is constructed.The alternating direction method of multipliers(ADMM)is used to solve the photoacoustic image model.Experimental results show that the proposed algorithm has better performance and visual effects,and can effectively preserve the detailed information of the original image.