Research On The Estimation Of Mean Transit Time In MRI Perfusion Imaging Based On Deconvolution Method

Magnetic resonance perfusion imaging is a kind of functional magnetic resonance imaging technology,which can reflect the micro vascular distribution and blood flow of the tissue,and can provide information of the anatomic structure,vascular imaging and organization function.The mean transit time(MTT)is one of the dynamic parameters which is very sensitive to distinct ischemia brain and normal one and is very important in the diagnosis of cerebral ischemia.In this thesis,the principle and application of magnetic resonance perfusion imaging in brain diseases and other diseases are reviewed firstly.And then,through the formula derivation for the basic principle of MRI cerebral blood perfusion imaging,a deconvolution method for estimating the MTT is put forward and the ideal dynamic functions are simulated.As direct deconvolution method is more complicated,so that MTT values are estimated by deconvolution using the methods of Tikhonov regularization based on weighted generalized cross validation(W-GCV),the total least squares regularization(TTLS)and the modified truncated singular value regularization(MTSVD).Meanwhile,the dwell function is simulated and the influence factors such as the singal noise ratio(SNR)and trace delay are also analyzed.The results show that the effect of SNR on estimated values of three kinds of regularization was gradually decreasing with increasing of SNR.The estimated values of MTT by the Tikhonov regularization based on W-GCV are closest to the reference values.As the increasing of the trace time delay,the MTT values of three regularization methods increase and the increased values are less than the delay value.Finally,pseduo color maps of the real cerebral perfusion imaging are produced by Tikhonov regularization based on W-GCV.The ischemia situations of the regions of interest in brain are analyzed with the estimated results combining with the cerebral blood flow(CBF)maps.The MTT ratio of the affected side and the healthy side is higher than1.63,which shows that the brain is in a state of ischemia.The CBF value of the affected side obtained from CBF pseudo color picture is in the range of 20~35ml/min,which indicates that N protein synthesis stops,and if the value of CBF doesn't decline continually,the tissue in brain will still be alive.The above results can provide the basis for clinical diagnosis of brain is ischemia.