MR-based thermometry of laser induced thermotherapy: Temperature accuracy and temporal resolution in vitro at 0.2 and 1.5 T magnetic field strengths

摘要

Purpose To evaluate MR-thermometry using fast MR sequences for laser induced interstitial thermotherapy (LITT) at 0.2 and 1.5 T systems. Methods & Materials In-vitro experiments were performed using Agarose gel mixture and lobes of porcine liver. MR-thermometry was performed by means of longitudinal relaxation time (T1) and proton resonance frequency shift (PRF) methods under acquisition of amplitude and phase shift images. Four different sequences were used for T1 thermometry: A gradient-echo (GRE), a True Fast Imaging with Steady Precession (TRUFI), a Saturation Recovery Turbo-FLASH (SRTF), and an Inversion Recovery Turbo-FLASH (IRTF) sequence (FLASHaFast Low Angle Shot). PRF was measured with four sequences: Two fast-spoiled GRE sequences (one as WIP sequence), a Turbo-FLASH (TFL) sequence (WIP sequence), and a multiecho-TrueFISP sequence. Temperature was controlled and verified using a fiber-optic Luxtron device. The temperature was correlated with the MR measurement. Results All sequences showed a good linear correlation R 2 = 0.97-0.99 between the measured temperature and the MR-thermometry measurements. The only exception was the TRUFI sequence in the Agarose phantom that showed a non-linear calibration curve R 2 = 0.39-0.67. At 1.5 T, the Agarose experiments revealed similar temperature accuracies of 4-6°C for all sequences excluding TRUFI. During experiments with the liver, the PRF sequences showed better performance than the T1, with accuracies of 5-12°C, contrary to the T1 sequences at 14-18°C. The accuracy of the Siemens PRF-FLASH sequence was 5.1°C. At 0.2 T, the Agarose experiments provided the highest accuracy of 3.3°C for PRF measurement. At the liver experiments the T1 sequences SRTF and FLASH revealed the best accuracies at 6.4 and 7.0°C. Conclusion The accuracy and speed of MR temperature measurements are sufficient for controlling the temperature-based tumor destruction. For 0.2 T systems SRTF and FLASH sequences are recommended. For 1.5 T systems SRTF and FLASH are the most accurate.