Research And Application Of Key Technologies For High-Performance Magnetic Resonance Imaging Spectrometer

To meet the urgent demand for high signal-to-noise ratio(SNR)magnetic resonance(MR)imaging in clinical diagnosis and fundamental brain science research,this dissertation focuses on the important component—spectrometer,which affects the SNR of MR imaging,as an impetus for enhancing the performance of magnetic resonance imaging(MRI)system.Three pieces of research were conducted based on spectrometer: highperformance analog-to-digital converter(ADC)application technology? dynamic gain workflow with a K-space splicing model? and reducing noise correlation between receiving channels.The main work of this dissertation is as follows:(1)Research I initiates with an investigation into the high-performance spectrometer hardware technology focusing on the ADC,a critical hardware component that affects the SNR of MR imaging.The experiment verified the impact of improving key ADC performance(including resolution accuracy and sampling rate,etc.)on imaging SNR.To address issues of the DC component and sampling clock jitter that significantly affect the ADC SNR,corresponding solutions were proposed and their effectiveness was verified.Consequently,a scheme of receiver hardware framework that both balances the performance of main hardware parameters and has the engineering feasibility was proposed.(2)The MR receive signals under dynamic gain control can take better advantage of the dynamic range of the receiver’s ADC,thereby obtaining a higher imaging SNR.Based on this,research II proposed integrating the K-space splicing model into the dynamic control process of the receive gain.Firstly,the data normalization method in the process of K-space splicing was studied.Then K-space splicing experiments were carried out along the phase and frequency encoding directions,respectively.Thus,a K-space splicing model that balances feasibility and enhanced SNR was proposed,and it was integrated into the dynamic gain workflow.Finally,dynamic adjustment of gain during a single scan and K-space splicing were implemented and validated to improve the imaging SNR,with the highest increase in SNR being 32.99%.(3)In response to the issue that noise correlation between channels can reduce the imaging SNR,research III found that the correlation coefficient between channels of the spectrometer receiver is superior to that of the receiving coil.Inspired by this,research III optimized the parallel acquisition process and verified that utilizing the remaining channels of the receiver to acquire noise data can enhance the reconstructed image SNR.It proposed a study integrating the method with noise pre-whitening,which proves to be advantageous through imaging.In summary,with the aim of enhancing the SNR of MRI,this dissertation systematically studies the key technologies of high-performance spectrometer from hardware technology research to workflow optimization.In terms of spectrometer hardware,a set of high SNR sampling and signal processing schemes with engineering applicability has been developed.Regarding the spectrometer workflow,a dynamic gain process incorporating the K-space splicing model and a parallel acquisition process that reduces noise correlation have been constructed.The research results indicate that the application of the key technologies of high-performance spectrometer studied in this dissertation to the current mainstream 1.5T/3T MRI systems can significantly improve the imaging SNR,which is of great significance for enhancing the accessibility of high-performance MRI equipment required for clinical diagnosis and brain science research.