SVD Compression for Magnetic Resonance Fingerprinting in the Time Domain

摘要: Magnetic resonance (MR) fingerprinting is a technique for acquiring and processing MR data that simultaneously provides quantitative maps of different tissue parameters through pattern recognition algorithm. A predefined dictionary models the possible signal evolutions simulated using Bloch equations with combinations various completed by computing inner product between observed each predicted signals within dictionary. Though this matching algorithm has been shown to accurately predict interest, one desires more efficient method obtain images. We propose compress singular value decomposition, which will provide low-rank approximation. By compressing size in time domain, we are able speed up algorithm, factor 3.4-4.8, without sacrificing high signal-to-noise ratio original scheme presented previously.