| Endoscopic photoacoustic tomography(EPAT)enables structural and functional imaging of cavity tissues for the diagnosis of coronary atherosclerotic diseases and digestive tract lesions.The limitation of laser illumination device,the complex optical properties of tissues and the attenuation of incident light in tissues will lead to the inhomogeneous distribution of optical fluence,thus resulting in the degradation of reconstructed image quality and the reduced imaging depth.This thesis focuses on the issues of the imaging accuracy degradation and the reduction in the image quality caused by the inhomogeneous optical fluence distribution in EPAT in order to improve the accuracy of optical absorption coefficient recovery.The main work involves two aspects.First,a method for reconstruction of optical absorption coefficient distribution map based on radiance Monte Carlo(RMC)model is proposed.The images representing the optical absorption coefficient distribution maps in the cross sections of the cavity are reconstructed by alternately optimizing two nonlinear least squares problems.Second,a method for simultaneous reconstruction of optical absorption coefficient and optical fluence distribution is proposed.The sparse representations of optical absorption coefficient and optical fluence are obtained by sparse decomposition of absorbed optical energy density with the greedy algorithm.Finally,the simultaneous reconstruction is achieved by sparse matrix decomposition.The validity of both methods has been verified by simulation data and phantom study.Comparison demonstrations are also conducted comparing the reconstructions with the proposed methods,two-step algorithm,one-step algorithm,and the iterative algorithm based on other light transport models.The results indicate that the proposed methods improve the reconstruction accuracy of the optical absorption coefficient and verify the superiority of the proposed method to other methods. |
PDF Full Text Request