An internal reference model-based PRF temperature mapping method with Cramer-Rao lower bound noise performance...

Area of Science:

• Medical Imaging
• Biophysics
• Magnetic Resonance Imaging

Background:

• Conventional MR thermometry methods face challenges including lipid proton interference, motion artifacts, and magnetic field drift.
• Accurate temperature monitoring is crucial for various medical applications, including thermal therapies and disease diagnosis.

Purpose of the Study:

• To develop and validate a novel model-based MR thermometry technique that utilizes a fat signal as an internal reference.
• To overcome the limitations of conventional phase difference methods in MR thermometry.

Main Methods:

• A multi-echo gradient echo (GRE) sequence was employed with a fat signal serving as an internal reference.
• The extended Prony and Levenberg-Marquardt algorithms were used to fit the signal model to water and fat signals, estimating temperature-dependent chemical shifts.
• Noise analysis was performed using the Cramer-Rao lower bound to assess algorithm performance and parameter influence.

Main Results:

• The model-based MR thermometry achieved a maximum temperature estimation error of 0.614°C with a standard deviation of 0.06°C when compared to thermocouple measurements.
• An optimal water:fat signal ratio of approximately 0.66:1 was identified for maximizing temperature estimation accuracy in phantom experiments.
• The method demonstrated feasibility and robustness in temperature mapping.

Conclusions:

• The presented model-based MR thermometry technique effectively addresses limitations of conventional methods, offering improved accuracy and reliability.
• This approach holds promise for precise, non-invasive temperature mapping in clinical and research settings.
• Further optimization of the water:fat signal ratio can enhance the precision of MR thermometry.