Regional Estimation of T1 Metabolite Relaxation using 2D MRSI and a Bootstrap Approach at 1.5 T
Noworolski S, Sdika M, Srinivasan R, Henry R, Ratiney H, Pelletier D
University of California
In clinical spectroscopic studies, the use of short repetition times (TR) enables one to meet time constraints but also requires knowledge of accurate metabolite longitudinal relaxation time values (T1) to correct for the resulting strongly T1-weighted signals. We propose a new method to estimate the metabolite T1, at 1.5T, using MR spectroscopic imaging (MRSI) data. The very low signal to noise ratio problem is tackled by using an approach combining Principal Component Analysis (PCA) denoising, bootstrap-like data averaging, and quantitation results quality assessment. A gray matter (GM) versus white matter (WM) contribution to the T1 estimation is also investigated.