MR Fingerprinting for Thermometry: A Comprehensive Literature Study
Summary
Magnetic Resonance Fingerprinting Thermometry (MRFT) is evolving as an innovative
and adaptable framework for non-invasive temperature mapping in thermal therapy.
Traditional MR thermometry methods, notably Proton Resonance Frequency Shift
(PRFS)-based methods, are still the preferred method for temperature monitoring in
aqueous tissues due to their high sensitivity and linear relationship with temperature.
PRFS approaches have drawbacks including motion sensitivity, magnetic field drift, and
low dependability in fat-rich or heterogeneous regions. MRFT, which takes advantage of
Magnetic Resonance Fingerprinting's (MRF) multi-parametric capabilities, offers a
viable solution to some of these limitations.
This review looks at numerous MRFT implementations, concentrating on their technical
designs, parameter sensitivity, and temperature estimating methods. MRFT's future
potential is stressed. Temperature could be modeled as a latent variable influencing T1,
T2, and Δf, resulting in extensive multi-parametric dictionaries. This would enable
temperature estimation in tissues where PRFS is insufficient. Furthermore, new
computational techniques, such as deep learning and partial volume mapping, can
speed up reconstruction and enhance accuracy in heterogeneous tissues.
MRFT offers a paradigm shift in MR thermometry since it combines signal modeling,
biophysics, and advanced computation. With further development and clinical validation,
it could provide a robust and precise method to real-time, tissue-specific heat monitoring
in a wide range of therapeutic situations.