INTRODUCTION

T₂ relaxometry can accurately quantify extravascular liquid accumulation. T₂ parametric maps could be biomarkers of infectious diseases having consequence in the brain. Computation of T₂ maps from multiple-echo images during a repetition time (TR) interval in conventional and fast spin-echo images has been used before [1-4]. Dual echo time (TE) T₂ FLAIR (Fluid-Attenuated Inversion Recovery) minimizes the partial volume effects from cerebrospinal fluid (CSF) and meets the acquisition time required in a BSL-4 laboratory, where cohorts of animals are imaged during the disease on a tight schedule until terminal stages [5]. However, the prescribed TE to compute T₂ varies among slices due to k-space mapping in fast spin-echo and an effective echo time (TE_eff) should be considered. We hypothesize that accurate TE_eff can be estimated by optimization based on semi-empirical values for each slice ordering. To evaluate this hypothesis, four gel phantoms were scanned and their T₂ values were computed using echo times around the prescribed TE. Dependence of computed T₂ in a healthy non-human primate on echo times was investigated.

METHODS

A dedicated Philips Achieva 3T MRI scanner (Philips Healthcare) was used for imaging. Four gel phantoms (Leeds Test Objects, North Yorkshire, UK) with reported T₂ values similar to those expected in the brain were used: T₂p₁ = (52+/-2) ms; T₂p₂ = (113+/-3) ms; T₂p₃ = (157+/-5) ms; T₂p₄ = (214+/-6) ms (T=19°C, 3% error). Dual TE T₂ FLAIR images (FA=90°; TR=10000 ms; TE_a=100-120 ms; TE_b=10-35 ms) with either linear or low/high slice ordering were acquired (Table 1) and processed with an in-house extension for MIM v.6.7 (MIM Software, Cleveland, OH, USA). T₂ maps were computed (T₂ = (TE_a – TE_b) / (ln(S_a)-ln(S_b)) using increased and decreased TE with respect to the prescribed values by a correction factor F(%). For each phantom, the mean T₂ within a volume of interest (VOI) was calculated, the total error between the computed T₂ and the T₂ reported by the manufacturer was minimized, and TE_eff=TE(1+F_opt/100) was calculated using the optimal factor F_opt. A healthy rhesus macaque was scanned to compare T₂ maps. The animal was housed in an Assessment and Accreditation of Laboratory Animal Care International accredited facility.

RESULTS

Figure 1 shows the setting for all phantoms in dual TE T₂ FLAIR images. All computed T₂ maps for different phantoms, slice ordering, and correction factors exhibited standard deviations within the VOIs comparable to the reported T₂ errors. Figure 2 shows the total percent error in all phantoms between the computed and reported T₂ relaxation times calculated as a function of F (Simulations #1-#4). The effective echo times TE_eff corresponding to the optimal correction factor F_opt for each slice ordering in the k-space are shown in Table 1. For the lowest short TE of 10 ms and all long echo times ranging from 100 ms to 115 ms the total percent error and the maximum error for an individual phantom remained below 1.9% and 3.5%, respectively. The performance of linear and low/high slice ordering schemes was compared for TE_a = 100 ms, TE_b = 20 ms and 35 ms, and the percent error was smaller for the linear ordering: 4.2% and 7.7% compared to 12.1% and 11.0%. Dual TE T₂ FLAIR images of the non-human primate (Figure 3) were used to compute T₂ maps using different echo times (Figure 4) around TE_a=100 ms (linear) and TE_b=35 ms (linear). A minimum total error of 7.7% with respect to all phantoms was achieved when the optimal correction factor F was used.

DISCUSSION

The brain is an organ of interest in the high-consequence infectious disease mechanics and therapy monitoring. Subtle BBB disruption and edema can be identified and quantified using T₂ maps [7]. However, accurate estimation of T₂ relaxation times is required to accurately correlate imaging finding with biological processes and therapeutic efficiency. The use of phantoms to estimate water content was proposed [8]. In this work, however, we present a methodology to improve the accuracy of T₂ estimation using fast spin echo sequences and phantoms. Reliable relaxation times can be obtained by adding a phantom-calibration step to the T₂ map calculation. Calibration is needed for further comparison to R₂* maps to compute of R₂’ maps as a biomarker of iron deposition, and reliable synthetic MR images based on T₁/T₂ maps.

Acknowledgements

The authors thank Laura Bollinger and Jiro Wada, IRF, for technical writing services and figure preparation and layout, respectively. This work was supported by NIAID Division of Intramural Research and NIAID DCR and was performed under Battelle Memorial Institute contract (No. HHSN272200700016I) with NIAID. Animals were housed in an AAALAC-International-accredited facility. All experimental procedures were approved by the NIAID Division of Clinical Research (DCR) Animal Care and Use Committee and were in compliance with the Animal Welfare Act regulations, Publish Health Service policy, and the Guide for the Care and Use of Laboratory Animals recommendations.