Study On Diffusion Correlation Tomography Of Blood Flow Based On The Nth-order Linear Algorithm

Before the tumor is formed,the tissue or organ already has abnormal metabolic function,presenting a state of increased blood flow and low blood oxygen in this tissue or organ.The near-infrared diffused light technology uses the blood flow and blood oxygen status inside the tissue reflected by the information carried by the escaped photons to infer the metabolic function of the tissue,and then predicts the pathological trend of the tissue.Compared with CT,MRI and other morphological images,Diffuse Correlation Tomography(DCT)can detect the changes of tissue hemodynamics functional earlier,and it has important clinical significance for predicting the occurrence and development of tumors.As an emerging functional imaging technology,DCT has attracted the research and development of many research institutions in the United States,Germany,Spain,Singapore and other countries.Among them,the analytical methods and finite element method(FEM)proposed by the Harvard University Massachusetts General Hospital,the University of Pennsylvania and the University of Kentucky were applied widely.However,The analytical method needs to assume that the tissue is regular and layered structure.The finite element method overcomes this shortcoming,but it still cannot make full use of the autocorrelation data.So they are more susceptible to noise.To adapt to any irregular organization and make full use of autocorrelation data,based on the newly proposed Nth-order linear algorithm(NthL algorithm),this paper conducts an in-depth study of DCT hardware equipment,imaging algorithms,data correction and calibration,and clinical applications.The main work is as follows:(1)The NthL algorithm proposed in this paper is a new algorithm independent of the analytical method and the finite element method.It is based on the diffusion equation.But with the help of the escape photon information given by Monte Carlo simulation,the slope of autocorrelation function can be obtained by its Taylor series expansion.Keeping the slope as the projection data in the inverse problem,the complex non-Linear problems are transformed into linear equations to be solved.Its unique advantages are: avoiding complex differential equation solving;realizing blood flow imaging of any tissue shape and structure;independently selecting the multi-delay data range on the autocorrelation function curve according to the noise level,to reducing noise interference;(2)The linear equations obtained from the NthL algorithm are seriously ill-conditioned.In order to optimize the solution and increase its stability,the TV sparse model as the regularization factor was embed into the split-Bregman reconstruction algorithm to form the Bregman-TV reconstruction algorithm.Through computer simulation,phantom experiments and clinical trials,It is proved that the algorithm can better keep the continuity and edge sharpness of the reconstructed image.(3)To reduce the equipment cost of NthL-DCT and increase the numbers of the source and detector(S-D)sensor array,the designed optical switch array multiplexes a group of S-D sensor arrays into multiple groups of S-D sensor arrays in a time-division multiplexing manner.Furthermore,with the goal of maximizing the detection area and depth,and ensuring the balance of voxel photons,the distribution modes of the S-D sensor array on some tissue models was designed.The first large ROI(Region of Interest)contact DCT equipment was designed;(4)To verify the sensitivity of the NthL-DCT technology to the changes of blood flow index(BFI)and simulate the clinical trial scenario,we designed a phantom experimental model with the same optical parameters but different morphological structures and controllable flow rates;Moreover,according to the data characteristics during the reconstruction process,we determined the NthL-DCT data screening,correction and calibration methods to overcome the influence of ambient light,vibration and other noise in phantom experiments and clinical trials;(5)Since the optical parameters of the tissue to be tested cannot be obtained in advance in clinic,we firstly derived the mathematical relationship between optical parameters and BFI.After verification of phantom experiments,It is believed that the clinically preset optical parameters,even the reduced scattering coefficients that have a large negative impact on the BFI,have no significant impact on the contrast of the reconstructed image.In other words,there is no significant impact on the disease diagnosis.In conclusion,clinical tests on animals and humans have proved that the new DCT equipment and method based on the NthL algorithm are more sensitive to the blood flow.And the diagnosis similarity with MRI is as high as 68.4%.So it can be used as a low-cost,portable,non-invasive new technology to realize the diagnosis and treatment of some diseases,such as,breast diseases,brain diseases(brain tumor,stroke,respiratory sleep syndrome,etc.)and skeletal muscle diseases.