Contrast Optimization And Signal Extraction Of Chemical Exchange Saturated Transfer MRI

Chemical exchange saturation transfer（CEST）imaging shows great potential in clinical applications.It can not only image chemical groups with very low concentration individually,but also detect the changes in the environment where the groups are located,so it is very suitable for molecular imaging.Clinical application of CEST mainly focuses on the brain,such as the diagnosing and grading of ischemic strokes,brain tumors,and other diseases.With the development of CEST,the application of body CEST has gradually come into people’s sight.However,CEST sequences are often difficult to obtain high contrast in a short time,and there are still many challenges in CEST reconstruction process,which limit its wide application in clinical practice.In this thesis,we did three parts of work on the above problems.1.Saturation strategy optimization and imaging sequence design in CESTClinical magnetic resonance imaging（MRI）needs to obtain good contrast of the pathological region within the appropriate time.The contrast of CEST imaging depends largely on the strategy of the saturation module and the design of the signal acquisition module.In this study,numerical simulation and experimental exploration were combined to obtain the optimal saturated module scheme for separated CEST imaging,and a fast CEST sequence with a special k-space filling strategy was designed.Compared with the traditional K space priority filling strategy,our results showed that a better contrast could be obtained.2.A CEST signal extraction method of brainThe post-processing method has always been a very important link in CEST imaging.There are many shortcomings in the current clinical CEST imaging,including the inability to obtain pure contrast of interested groups while using the traditional asymmetric analysis method.In this study,we proposed a novel method based on a neural network fitting the background reference Z spectrum to extract the brain CEST signal and obtained pure amide proton transfer（APT）and nuclear Overhauser enhancement（NOE）imagings.3.A post-processing method of body CEST imagingBody CEST imaging is often affected by large magnetic field offset and unstable fat suppress effect,which limits its clinical application.In this study,we proposed a novel CEST post-processing method based on machine learning.Through recognizing the characteristics of the acquired Z-spectrum,the background reference Z-spectrum and the B₀ offset were obtained and used to correct the acquired Z-spectrum without the need for additional sequences.The APT effect and NOE effect could be calculated by subtracting the background reference Z-spectrum from the acquired Z-spectrum and theoretically not affected by the effect of fat suppress.