Repeatability and reproducibility of metabolite T2 estimates assessed with 1D and 2D models

Motivation: Metabolite T2 estimations are critical for quantifying concentrations and may provide biomarkers in their own right. No studies have characterized their reliability. Goal(s): To establish test-retest repeatability and reproducibility of T2 estimation and investigate whether multi-spectrum modeling improves them. Approach: We acquired repeated acquisitions of multi-TE sLASER —two sessions with two back-to-back repeats per session —in the CSO and PCC. Then we estimated metabolite T2s using 1D and 2D modeling approaches. Results: Within- and between-session reproducibility of major-metabolite (tNAA, tCr, tCho) T2s were below 5% for 1D models. T2 estimates from multi-spectrum models deviated somewhat but were largely comparable. Impact: This study of metabolite T2 reproducibility is the first of its kind and suggests metabolite T2 estimates are highly reproducible. Our early implementation of the 2D multi-TE model proved promising with scope for further development.