EPI proton resonant frequency temperature mapping at 0.5T in the brain: Comparison to single-echo gradient recalled echo.

Abstract

Purpose: Evaluate the use of both single-echo gradient recalled echo (SE-GRE) and EPI approaches to creating temperature maps on a mid-field head-only scanner, both in vivo and on a tissue mimicking gel.

Methods: Three 2D protocols were investigated (an SE-GRE, single-shot EPI, and an averaged single-shot EPI). The protocols used either a gradient recalled acquisition or an echo planar acquisition, with EPI parameters optimized for the longer $T_2^{*}$ at lower field-strengths. Phantom experiments were conducted to evaluate temperature tracking while cooling, comparing protocol to measurements from an optical fiber thermometer. Studies were performed on a 0.5T head only MR scanner. Temperature stability maps were produced in vivo for the various protocols to evaluate precision.

Results: The use of an EPI protocol for thermometry improved temperature precision in a temperature control phantom and provided an 18% improvement in temperature measurement precision in vivo. Temperature tracking using a fast (<2 s) update rate EPI thermometry sequence provided a similar precision to the slower SE-GRE protocol.

Conclusion: While SE-GRE PRF thermometry shows good performance, EPI methods offer improved tracking precision or update rate, making them a better option for thermometry in the brain at mid-field.