Font Size: a A A

PRF Based MR Thermometry At Low Field

Posted on:2019-08-19Degree:MasterType:Thesis
Country:ChinaCandidate:Y P ChenFull Text:PDF
GTID:2394330542499261Subject:Biomedical engineering
Abstract/Summary:PDF Full Text Request
Noninvasive magnetic resonance thermometry(MRT)using proton resonance frequency(PRF)is a promising technique for monitoring and guiding tissue temperature changes during tumor thermal therapy.Conventional PRFs can be used to determine tissue temperature differences from the phase differences of gradient recalled echo(GRE)or spoiled-gradient echo(SPGR)sequences.However,because baseline phase information is required before obtaining the phase difference information,the conventional PRF method is motion sensitive.External magnetic field drift is another drawback of the PRF method for long-term interventional ablation therapy.Low-field MR scanners with open configuration are more suitable for interventional procedures than closed systems.However,the magnetic field drift should be much more in low-field permanent magnet than high-field superconducting magnet.With the intention of overcoming the influence of external magnetic field drift,RF method with first-order polynomial fitting method was proposed for the first time to investigate the feasibility and accuracy of quantitative MR thermography during hyperthermia procedures.Compared with the high-order polynomials fitting,the PRF method using first-order polynomial fitting could achieve the smallest temperature errors in unheated phantom,ex vivo porcine liver and in vivo human brain experiments.In the ex vivo porcine liver microwave ablation procedure,the temperature errors between MRT and fiber optic probe of all but six temperature points were less than 2 ?.Moreover,to overcome the influences of motion,this study proposed a minimum curvature surface fitting(MCSF)modelling-based referenceless PRF method that considers distance information between pixels within the region of interest(ROI)for temperature mapping with low-field scanners.All experimental results showed that temperature root mean square error(RMSE)calculated by the MCSF model were lower than those computed by the conventional PRFs using high-order polynomial fitting.For ablation experiments,the temperature errors between the fiber optic probe and the proposed method of 42 temperature points were all less than 2 ?.MCSF modelling-based referenceless PRFs method has the potential to become a reliable and easy-to-use tool for monitoring tissue temperature changes at 0.35T.
Keywords/Search Tags:Low field, Temperature mapping, Proton resonance frequency, Polynomial model, Minimum Curvature Surface Fitting
PDF Full Text Request
Related items